Friday 28 September

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
08:30 - 10:00

Plenary Session 4
Movement Disorders

Moderators: Miroslav GALANDA (Kosice, SLOVAKIA), Alex GREEN (Consultant Neurosurgeon) (Oxford, UK), Michael SCHULDER (Vice Chair, Neurosurgery) (Lake Success, NY, USA)
08:30 - 08:40 #14471 - O092 Resting-state fMRI relates to clinical response after Vim radiosurgery for essential tremor.
O092 Resting-state fMRI relates to clinical response after Vim radiosurgery for essential tremor.

Introduction: Essential tremor (ET) is the most common movement disorder. Drug-resistant ET can benefit from standard procedures (deep brain stimulation, thalamotomy) or minimally invasive high-intensity focused ultrasound (HIFU) or ventro-intermediate nucleus (Vim) radiosurgery (RS). Resting state fMRI (rs-fMRI) is a non-invasive imaging method acquired in absence of a task. We examined whether rs-fMRI correlates with tremor score on the treated hand (TSTH) improvement 1 year after Vim RS.

Methods: We included 17 consecutive patients treated with left unilateral Vim RS in Marseille, France. Tremor score evaluation and rs-fMRI were acquired at baseline and 1 year after Vim RS. Resting-state data (34 scans) were analyzed without a priori hypothesis, in Lausanne, Switzerland. We used data-driven multivariate analysis (i.e., independent component analysis (Calhoun et al., 2001; Beckmann et al., 2005) to conduct whole-brain analysis without prior assumptions. Statistical investigation was implemented in Statistical Parametric Mapping (London, UK; version 12) as an analysis of variance (Anova) flexible factorial model, on each separate component, by using individual subject-level maps, to take into account time point (pretherapeutic versus 1 year after Vim RS), clinical response (less or equal versus more than 50% improvement of the TSTH), as well as the interactions between. Bonferoni correction was used to deal with number of models (20). We then reported corrected p-values using conventional cluster-level family wise error (FWE) correction. For relevant interconnectivity values, there was no influence of age or disease duration (p>0.05). Based on degree of improvement in TSTH, to consider Vim RS at least as effective as medication, we separated two groups: 1, <= 50% (n=6, 35.3%); 2, > 50% (n=11, 64.7%). They did not differ statistically by age (p=0.86), duration of symptoms (p=0.41) or MR signature volume at 1 year (p=0.06).

Results: We report TSTH improvement correlated with interconnectivity strength between salience network with left claustrum and putamen, as well as between bilateral motor network, frontal eye-fields and left cerebellum lobule VI with right visual association area (pFWE= 0.001; the former also correlated with lesion volume). For this former network, both pretherapeutic interconnectivity, as well as difference between 1 year and baseline, related to TSTH improvement after Vim RS. Furthermore, patients who alleviated less presented negative pretherapeutic interconnectivity (which increased to median positive values one year later), while those who alleviated more, had already positive pretherapeutic values (which decreased to a median of zero one year later) Longitudinal changes in time, between pretherapeutic state and 1 year later, showed additional associations in inter-connectivity strength between right dorsal attention network with ventro-lateral prefrontal cortex and salience network with fusiform gyrus (pFWE<0.001). At the opposite with the results described in the previous visuo-motor network, overall interconnectivity values decreased from slightly positive to the opposite slightly negative values 1 year after Vim RS. Furthermore, patients who alleviated less presented slightly positive pretherapeutic interconnectivity (which decreased to median of slightly negative symmetric, close to zero, one year later), while those who alleviated more, had pretherapeutic negative median values (close to zero, which slightly increased to a median of zero one year later).

Conclusions: Brain interconnectivity measured by resting-state fMRI relates to clinical response after Vim RS. There is a widespread visual network, which responds to Vim RS. Vim RS seems to bring interconnectivity in the visual areas back to normal for all patients, but the ones who had this region more functionally integrated pretherapeutically had much larger benefit. Beside this visual network, a major role is played by motor and attention systems. Inter-connectivity between visual and motor areas is a novel finding, revealing implication in movement sensory guidance. Whether the visual areas should be involved in the surgical targeting in the near future remains an open question. 

08:40 - 08:50 #16190 - O093 A prospective trial of MRIgFUS thalamotomy for ET: 2 year follow-up results.
O093 A prospective trial of MRIgFUS thalamotomy for ET: 2 year follow-up results.

Background: Magnetic resonance (MR) guided focused ultrasound (MRgFUS) has recently been demonstrated to be a safe and effective treatment for patients with medication refractory essential tremor (ET), but the long term durability of the procedure has not yet been evaluated. This study reports the results of MRgFUS thalamotomy for ET at the 2- year follow-up.

Methods: A total of 76 patients with moderate-to-severe ET, who had not responded to at least two trials of medical therapy, were enrolled in the original randomized study of unilateral thalamotomy (NEJM 375(8):730-9,2016) and evaluated using the clinical rating scale for tremor (CRST). Sixty-three of the patients continued in the open-label extension phase of the study with monitoring for 2 years.

Findings: Mean hand tremor scores improved by 53% at 1 year and by 55% at 2 years (from 19·8±4·9 to 9·0±5·1, mean change in CRST tremor score from baseline to 2 years, 10·8 points; 95% confidence interval, 7·7 to 10·2; p<0·001). Furthermore, the CRST disability score was improved by 64% at 1 year and by 60% at 2 years (from 16·4±4.5 to 6·5±5·0, mean change in the score from baseline to 2 years, 9·9 points; 95% confidence interval, 5·3 to 7·7; p<0·001). Two adverse events that occurred at the time of the original treatment resolved. There were no new delayed complications at 2 years.

Conclusion: MRIgFUS thalamotomy provides a significant and sustained improvement in contralateral tremor for patients with ET and improves their overall quality of life. Latent or delayed complications did not develop after surgery.

Jin CHANG, Rees COSGROVE (Boston, USA), Chang Kyu PARK, Nir LIPSMAN, Michael SCHWARTZ, Pejman GHANOUNI, Jaimie HENDERSON, Ryder GWINN, Travis TIERNEY, Takaomi TAIRA, Andres LOZANO, Howard EISENBERG, Jeff ELIAS
09:10 - 09:20 #14799 - O096 Optimizing stereotactic coordinates of Prelemniscal Radiations as target for the treatment of Parkinson’s disease. II. Individual variations in stereotactic location of fiber components: A probabilistic tractography study.
O096 Optimizing stereotactic coordinates of Prelemniscal Radiations as target for the treatment of Parkinson’s disease. II. Individual variations in stereotactic location of fiber components: A probabilistic tractography study.

Objective. To determine individual variations of Prelemniscal radiations fiber components as target to treat motor symptoms of Parkinson’s disease.

Material and Methods.Fiber composition of Raprl was determined in the two hemispheres of a group of 15 PD patients and 15 controls paired in sex and age, for a total of 60 hemispheres, using 3Tesla Magnetic Resonance imaging (MRI) and probabilistic tractography, Diffusion Weighted Images (DWI) with high angular resolution and constrained spherical deconvolution (CSD). Stereotactic position, regarding AC-PC length and level and midsagittal plane, of fiber tracts for “x”, “y” and “z” was evaluated regarding right and left hemispheres in the same person and among individuals. 

Results.Three fiber components of Raprl were identified with different connectivity: cerebellar-thalamic-cortical, Globus pallidum-peduncle-pontine nucleus and Orbital frontal cortex-mesencephalic. Fiber stereotactic position did not vary between right and left hemisphere in the same person. In contrast, variations of HPC location were significant for all tracts among subjects and more prominent for the orbitofrontal-mesencephalic tract. Such variations can be only determined by probabilistic tractography that could be used for planning electrodes’ trajectories in the future.

Conclusion.Individual optimum target to approach Raprl is better defined by probabilistic tractography. Tractography provides a platform for planning the stereotactic approach and conform volumes for DBS and lesions.

09:20 - 09:30 #16170 - O097 INTREPID: a prospective, double blinded, multicenter randomized controlled trial evaluating deep brain stimulation with a new multiple source, constant current rechargeable system in parkinson’s disease.
O097 INTREPID: a prospective, double blinded, multicenter randomized controlled trial evaluating deep brain stimulation with a new multiple source, constant current rechargeable system in parkinson’s disease.

Objective: INTREPID is designed to assess the improvement in motor function and quality of life in patients with Parkinson's disease (PD) following Deep Brain Stimulation (DBS) using a new device with multiple independent current sources that allowed for selective activation of individual contacts on the DBS lead thereby permitting a defined distribution of applied current. 365 characters 

Background: Deep Brain Stimulation (DBS) is an effective treatment for the motor signs and fluctuations associated with Parkinson's disease (PD). Although DBS efficacy has been substantiated by several randomized controlled trials (RCT), the degree of improvement varies significantly. The INTREPID Trial assessed improvement in motor function and quality of life in PD patients following bilateral subthalamic nucleus (STN) DBS using a new device with multiple independent current sources that allowed for selective activation of individual contacts on the DBS lead thereby permitting a defined distribution of applied current. 

Methods: INTREPID is a multicenter, prospective, double blinded RCT sponsored by Boston Scientific. Subjects were implanted bilaterally in the STN with a multiple source constant current DBS System (Vercise System). Blinded subjects were randomized to either receive active vs. control settings for a 12 week period. All assessments were completed by a blinded assessor. Following the blinded period, subjects received their best therapeutic settings. Improvement in motor function and quality of life was evaluated using PD diary, UPDRS, PDQ- 39, and a battery of neuropsychological assessments. Adverse events were recorded. 

Results: The study successfully met the primary endpoint (p < 0.001) with a mean difference of 3.03 ± 4.2 hours from baseline to 12 weeks between the active and control groups in ON time (PD diary), with no increase in antiparkinsonian medications. The study also met several of the secondary endpoints. The incidence of infection was 2.7% and peri-operative intracranial hemorrhage was 1%. 

Conclusions: The results of the INTREPID Trial demonstrate that the use of a multiple source, constant-current DBS System is safe and effective in the treatment of Parkinson's disease symptoms. 

Jerrold VITEK (Minneapolis, USA), Intrepid STUDY GROUP, Roshini JAIN, Lilly CHEN, Philip A. STARR
09:30 - 09:40 #16307 - O098 Has deep brain stimulation changed the natural history of Parkinson’s disease? A case control longitudinal study.
O098 Has deep brain stimulation changed the natural history of Parkinson’s disease? A case control longitudinal study.


Deep brain stimulation (DBS) is often regarded as the second therapeutic breakthrough after L-Dopa in the history of Parkinson’s disease (PD) therapies. However, the impact of DBS on the long-term course of PD has not yet been evaluated in a controlled manner.



We collected retrospective information on key disease milestones (recurrent falls, psychosis, dementia, and nursing-home placement) and death from clinical notes of PD patients treated with chronic subthalamic DBS >10 years (1999–2007) at our centre. A control group of PD patients similar in age at onset and age at baseline was extracted from a registry study (EuroPa) performed in 2003/2004 with corresponding retrospective data collection on long-term outcomes. Cox regression models were used to calculate hazard ratios (HR), adjusted for potential confounding variables.



Fifty-four patients with DBS and 54 patients without DBS at baseline were included. Groups were not significantly different with respect to age at onset, age at baseline, sex-distribution, and number of comorbidities at baseline. Compared to patients without DBS, patients treated with DBS were at lower risk of recurrent falls (HR=0.6; p=0.035) and of psychosis (HR=0.4; p=0.031). There was no significant difference in risk for dementia (HR=1.2, p=0.67), nursing home placement (HR=0.6; p=0.26), or death (HR=1.1; p=0.73).



Treatment with chronic subthalamic DBS was associated with longer intervals to recurrent falls and onset of psychotic symptoms. There was no evidence for beneficial effects of DBS on the long-term evolution of dementia, need for nursing home placement, or on overall survival.

Philipp MAHLKNECHT (Innsbruck, AUSTRIA), Marina PEBALL, Katherina MAIR, Mario WERKMANN, Michael NOCKER, Elisabeth WOLF, Wilhelm EISNER, Cecilia PERALTA, Sabine ESCHLBÖCK, Gregor WENNING, Peter WILLEIT, Klaus SEPPI, Werner POEWE
09:40 - 09:50 #16374 - O099 Deep brain stimulation for Parkinson’s Disease: refining the optimal location within the subthalamic nucleus.
O099 Deep brain stimulation for Parkinson’s Disease: refining the optimal location within the subthalamic nucleus.

Background: Individual motor improvement after deep brain stimulation (DBS) of the subthalamic nucleus (STN) for Parkinson’s disease varies considerably. Recently the medial STN border proved superior compared to MCP as anatomical reference for correlation of DBS location and motor improvement, and enabled defining an optimal DBS location within the nucleus. In order to confirm the superiority of the medial STN border as anatomical reference we again evaluated this new anatomical reference in a recent cohort of patients at our institution.

Methods: Motor improvement after six months of 74 STN DBS electrodes was categorized into non-responding, responding and optimally responding body-sides. Stereotactic coordinates of optimal electrode contacts relative to both medial STN border and MCP served to define theoretic DBS ‘hotspots’.

Results: Using the medial STN border as reference, significant negative correlation (Pearson -0.27, p < 0.02)  was found between the Euclidean distance from the centre of stimulation to this DBS hotspot and motor improvement. This hotspot was located at 2.4 mm lateral, 0.7 mm anterior and 1.5 mm superior relative to the medial STN border. Using MCP as reference, no correlation was found.

Conclusion: This is a second, larger, cohort of Parkinson’s disease who underwent STN DBS in which the medial STN border proved superior compared to MCP as anatomical reference for correlation of DBS location and motor improvement. The percentage of optimally responding body-sides (39%) in current cohort enabled further refinement of the optimal DBS location. Implementing the optimal point of location in our DBS workflow has possibly contributed to more optimal, and thus less variable, motor improvement for individual PD patients following STN DBS.

09:50 - 10:00 Discussion.

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
10:00 - 10:30



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
10:30 - 12:00

Plenary Session 5
Surgery for Depression

Moderators: Benjamin GREENBERG (USA), Marwan HARIZ (London, UK), Andres LOZANO (Chairman of Neurosurgery, University of Toronto) (Toronto, CANADA)
10:30 - 10:50 Genetic and Functional Architecture of Mood Disorders. Andrew MCINTOSH (Professor of Biological Psychiatry) (Edinburgh, UK)
10:50 - 11:10 DBS for depression. Volker COENEN (Head of Department) (Freiburg, GERMANY)
11:10 - 11:30 Stereotactic ablation (capsulotomy & cingulotomy) for depression. David CHRISTMAS (Consultant Psychiatrist) (Dundee, Scotland, UK)
11:30 - 11:50 The psychiatrist perspective on surgical results for depression. Keith MATTHEWS (Professor) (Dundee, UK)
11:50 - 12:00 #16347 - O100 Retrospective and prospective evaluation of tractography for bilateral anterior capsulotomy in patients with depression.
O100 Retrospective and prospective evaluation of tractography for bilateral anterior capsulotomy in patients with depression.

Introduction: Bilateral anterior capsulotomy (BAC) is an effective surgical procedure for patients with treatment-resistant major depression. The anterior limb of the internal capsule (ALIC) carries circuits associated with emotion and neurocognition. We analyzed the connectivity of the BAC lesions to identify ‘fingerprints’ associated with clinical outcomes. We also analyzed the feasibility of tractography to guide the selective ablation (or preservation) of pathways within the ALIC.

Methods: Ten patients were retrospectively analyzed following BAC surgery. These patients were divided into ‘responders’ or ‘partial responders’ based on the Beck Depression Inventory (BDI) score at one-year follow-up. These patients were matched with ten subjects obtained from a neuroimaging sample connectome. The lesions were segmented and transferred to the native space of the matched subjects to generate group-averaged probabilistic ‘fingerprints’ associated with a given outcome. We also generated the major fibers going through the ALIC (mesocorticolimbic , anterior thalamic radiations [ATR], limbic, and associative) and analyzed if the overlap of the lesions with either of these pathways were associated with outcome. Two patients, one patient with TRD and the other with treatment resistant obsessive compulsive disorder (TROCD) and co-morbid depression were treated with tractography guided capsulotomy of the limbic pathways as a proof of concept analysis.

Results: Six patients were responders (> 50% improvement in BDI) and four patients were partial responders (25-50% improvement). The responder map showed significant connections with limbic areas including ventromedial prefrontal cortex, anterior cingulate cortex, and lateral orbitofrontal cortex. The partial responder map showed significant connections with the same limbic areas and also significant stronger connectivity to associative areas including the dorsolateral prefrontal cortex, ventrolateral prefrontal cortex, and lateral orbitofrontal cortex. The overlap analysis showed that in the responder group, the involvement of the associative pathways was significantly less than the limbic pathways (p=.00). Conversely, in the partial responder group, there was no significant difference between the involvement of these pathways (p=.16). Finally, there was no significant difference in the involvement of the mesocorticolimbic tracts compared to the ATR in the two outcome groups (p=.17, p=.47). The two patients treated with tractography-guided ablation of the limbic pathways were responders after six months and no neurocognitive side effects were present after surgery. The tractography analysis showed the preservation of the associative pathways and interruption of the limbic pathways.

Conclusions: The optimum outcome following BAC surgery in this cohort was associated with interruption of key limbic areas and the relative preservation of associative pathways. Tractography is able to show patient-specific regional difference between associative and limbic pathways within the ALIC. This information is useful to identify the pathways that need to be destroyed or preserved during capsulotomy. Tractography-guided limbic capsulotomy has demonstrated to be safe and feasible and will be evaluated for long term results in patients with depression.  

Josue AVECILLAS-CHASIN, Trevor HURWITZ, Christopher HONEY (Vancouver, CANADA)

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
12:00 - 13:30



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
12:00 - 13:30



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
12:00 - 13:30


Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
12:00 - 13:30



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
13:30 - 14:35

Plenary Session 6
Pain Surgery

Moderators: Jonathan HYAM (Consultant) (London, UK), Roman LISCAK (head) (PRAGUE, CZECH REPUBLIC), Angelo FRANZINI (MILANO, ITALY)
13:30 - 13:45 Cluster Headache & Migraine. Angelo FRANZINI (MILANO, ITALY)
13:45 - 14:00 Cortical stimulation. Patrick MERTENS (Head of the department) (LYON, FRANCE), Andrei BRINZEU (MD) (Lyon, FRANCE)
14:00 - 14:15 Update on neurosurgery for chronic pain. Tipu AZIZ (Professor) (Oxford, UK)
14:15 - 14:25 #16338 - O101 Long-term Experience with Occipital and Supraorbital Nerve Stimulation for the Various Headache Disorders – an Institutional Case Series.
O101 Long-term Experience with Occipital and Supraorbital Nerve Stimulation for the Various Headache Disorders – an Institutional Case Series.

Long-term Experience with Occipital and Supraorbital Nerve Stimulation for the Various Headache Disorders – an Institutional Case Series.


Mahmoud Abdallat1,3, MD

Holger Joswig1, MD, FMH

Vahagn Karapetyan1, MD

Keith W. MacDougall1, MD, FRCSC

Paul E. Cooper2, MD, FRCPC

Andrew G. Parrent1, MD, FRCSC


1Department of Clinical Neurological Sciences, Division of Neurosurgery, London Health Sciences Centre, University Hospital, London ON, Canada

2Department of Clinical Neurological Sciences, Division of Neurology, London Health Sciences Centre, University Hospital, London ON, Canada

3Hannover Medical School, Department of Neurosurgery, Hannover, Germany 

Objectives: To assess whether occipital and supraorbital nerve stimulation (ONS; SONS) for the various headache disorders results in clinically meaningful long-term pain alleviation.

Materials and Methods:Retrospective chart analysis of a cohort of 90 patients (age 47.2±12.3 years, 57.8% female) suffering from migraine, cervicogenic headache, cluster headache, neuropathic pain of the scalp, tension-type headache and new daily persistent headache, with a pain course of headache of 13.4±14.1 years and a HIT-6 score of 64.4±7.6 undergoing ONS (58.9%), SONS (12.2%), or a combined ONS + SONS (28.9%) stage 1 (trial) and stage 2 (definite implantation) between 2007 and 2017. Relative change in visual analog scale (VAS) pain was displayed graphically over time and significances tested using the t-test.

Results:57 out of 90 patients (63.3%) were treatment-responders to a stage 1 trial lasting 22.1±9.4 days, with reduction of their average VAS pain to 35.8±24.8% of baseline (non-responders 99.1±24.1% during the trial; p<0.01). Stage 1 percutaneous lead implantation took 27±15.8 minutes and exposed the patients to 235.1±188.1 rad*cm2. Following stage 2 (75.8±33.8 minutes surgery time; 170±116.8 rad*cm2radiation exposure), average VAS pain remained <50% of baseline on long-term follow-up for up to 10 years. 2 patients (3.5%) requested hardware explantation because of lack of treatment effect. Stage 2 complications were 1 infection (1.8%) and 4 electrode dislocations (7%).

Conclusions: After careful patient selection, based on a positive response to a stage 1 trial ONS / SONS, a clinically meaningful long-term benefit can be achieved in chronic headaches patients who had failed medical management.

Mahmoud ABDALLAT (Amman, JORDAN)
14:25 - 14:35 #14474 - O101b Trigeminal ultrasonic nucleotractotomy for treatment of deafferentation facial pain.
O101b Trigeminal ultrasonic nucleotractotomy for treatment of deafferentation facial pain.

Introduction. The deafferentation facial pain syndrome, caused by various lesions of the trigeminal sensory root is a severe pathological condition leading to patients’ disability if not controlled. Many ablative procedures at the level of trigeminal descending tract and spinal trigeminal nucleus developed for the treatment of neuropathic facial pain including open tractotomy, percutaneous stereotactic nucleotractotomy, open caudalis dorsal root entry zone lesions and pontine stereotactic trigeminal nucleotractotomy. The major indications to trigeminal nucleotractotomy at this moment are following: tumours and trauma of peripheral branches of V CN, tumours and trauma of the root of V CN (anesthesia dolorosa), post-herpetic neuralgia, cluster headache and sometimes – failed surgery for trigeminal neuralgia and glossopharyngeal neuralgia. We present our results of ultrasonic trigeminal nucleotractotomy in 50 patients suffering from deafferentation facial pain.

Method. 50 patients suffering from neuropathic facial pain underwent ultrasonic trigeminal nucleotractotomy between the 1987 – 2018.  33 patients with deafferentation pain due to various mechanical lesion of trigeminal nerve or herpes infection, 14 with migraine-induced neuralgia and 3 with trigeminal neuralgia due to multiple sclerosis. All patients experienced the painful dysaesthesias and intractable pain located in the face and oral cavity. The sensation was described as a throbbing, burning pain in half of the face with anesthesia at the same side with no effect from medication. All procedures were done in the sitting position under the general anesthesia. After small unilateral occipital craniectomy and CI hemilaminectomy on the pain side dura was cut. Vertical lesion of the trigeminal nucleus caudalis and spinal trigeminal tract performed by ultrasonic microsurgical needle along the line between the C 2 dorsal root zone and the point 2 mm dorsally to the lowest rootlet of the vagus nerve, preserving all small vessels along the way (Fig.1). The level of lesion was extended above the obex if pain was located in medial part of face, lower part of face or in oral cavity.

Results. Good immediate results (relief of baseline pain or mild pain requiring no medical therapy) achieved in 48 (96%) patients and poor results in 2 (4%). The usual side-effect – ipsilateral ataxia more prominent in upper limb due to disruption of spinocerebellar tract and posterior column was noted in 12 patients. In vast majority of cases the ataxia was well tolerated and disappeared in few weeks.

Ipsilateral paresis and contralateral hypoesthesia because of damage of corticospinal and spinothalamic tracts was seen in 3 patients. No other neurological symptoms or mortality was seen. Late follow up was available in 40 patients. Good, fair (pain control with medication), and poor results (no control of pain) were preserved in 26, 4, and 10 patients retrospectively.

Conclusions.Trigeminal ultrasonic nucleotractotomy is effective and relatively safe in relieving intractable facial pain in highly selected patients.



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
14:30 - 15:00



Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
15:00 - 17:00

Parallel Session 11
Advance Imaging in Functional NS benefit & limitations of each

Moderators: Alexandre CAMPOS (Neurosurgeon - Consultant) (Lisbon, PORTUGAL), Julian EVANS (Manchester, UK), Sameer SHETH (Associate Professor of Neurosurgery) (Houston, USA)
15:00 - 15:15 Tractography for Capsulotomy. Sameer SHETH (Associate Professor of Neurosurgery) (Houston, USA)
15:15 - 15:30 Structural connectivity. Harith AKRAM (Neurosurgeon) (London, UK)
15:45 - 15:55 #16311 - O102 Holographic visualization of deep brain stimulation.
O102 Holographic visualization of deep brain stimulation.

Stereotactic neurosurgery for the placement of deep brain stimulation (DBS) electrodes is an inherently complex 3D problem.  Unfortunately, traditional tools to visualize stereotactic frame adjustments on the electrode trajectory or the electrode placement within the patient-specific neuroanatomy typically rely on 2D computer screens that require substantial degrees of imagination to resolve into a 3D understanding.  Neurosurgeons currently overcome these limitations via experience and mental visualization; however, augmented reality (a.k.a. mixed reality) provides a new medium to work with patient-specific data in a context that directly mimics the surgical environment.  Therefore, we adapted our academic neurosurgical navigation software tool, StimVision, to run within the Microsoft HoloLens platform.  Importantly, this is not virtual reality (VR), where the user is isolated in a completely artificial space.  This is augmented reality (AR), where the holograms are visible in addition to the real world and real people surrounding the model.  Our StimVision HoloLens application is capable of reading in patient imaging data, defining the stereotactic coordinate system, visualizing adjustments to the frame system, enabling interactive positioning of DBS electrodes, integrating tractographic representations of axonal pathways, and predicting the direct activation of the axonal pathways as a function of the stimulation parameter settings.  This final step is accomplished with simulations that couple DBS electric field calculations with multi-compartment cable models to predict action potential generation in the various axonal pathways as a function of the patient-specific electrode location (arc angle, ring angle, X, Y, Z position) and stimulation parameter settings (contact, pulse width, amplitude, frequency).  As such, StimVision HoloLens provides an AR DBS platform that improves the visualization and quantification of clinical therapy, all within a novel teaching environment for neurosurgical training.

Mikkel PETERSEN, Angela NOECKER, Jeff MLAKER, Mark GRISWOLD, Cameron MCINTYRE (Cleveland, USA)
15:55 - 16:08 #16319 - O103 Geometric distortions in stereotactic MR brain imaging: acquisition center sensibility and constructor distortion filtering.
O103 Geometric distortions in stereotactic MR brain imaging: acquisition center sensibility and constructor distortion filtering.

1. Introduction

Stereotactic surgery is a well-established treatment approach for both Deep Brain Stimulation (DBS) and Gamma Knife radiosurgery (GK RS). MRI-based targeting in DBS and in GK RS is a critical issue, as it aims at both providing convenient contrast to see the targets and insure non-distorted images in a clinically acceptable time. Thus, evaluation of geometric distortions in clinical MRI systems is crucial for pre-operative targeting planning.

Sources of distortions can be due to the magnet itself, to the MR sequence or to the patient. Geometric distortions come from different factors: 1) gradient non-linearity (magnet), 2) inhomogeneity of magnetic field (magnet and sequence) 3) magnetic susceptibility and chemical shift (sequence and patient). Several aspects have already been well studied in the litterature [1,2,3], but some work remains to be done.

In this study, we will use a geometric phantom and we will focus on 3D anatomical T1-weighted images used for targeting on 1.5T GE and 3T Siemens magnets. Our aim will be to evaluate how the laser positioning (used to center the object in the tunnel) influences the distortion field. In practice, we will study the distortion field as a distance to the magnet center by changing the laser position on the head coil, and evaluate the errors.

And as MR constructors include distortions-correction filters to reduce gradient non-linearities on a number of sequences, we will also evaluate these errors on corrected images, allowing  to quantify the effect of these filters.

2.  Material and Methods

We used the GRID 3D phantom (MODUS QA Inc.), a rectilinear grid (10 mm spacing) of 140 x 130 x 110 mm3, filled with acrylic plates, a solution of copper sulphate and deionized water.

We acquired FSPGR (GE Optima MR 450W, 1.5T) and MP2RAGE (SIEMENS Skyra, 3T) images, twice. First, the laser was centered relative to the phantom, second it was positioned on the superior part (approximately 55 mm) of it .The acquisitions were done with and without distortions-correction filters provided by GE and Siemens (see Table 1).

An in-house Python-based module was developed and implemented in 3D Slicer [4] for the automatic assessment of geometric distortions with the phantom. All images were preprocessed by cropping, denoising [5] and resampling to 0.5 mm iso-voxel.

Detection of the intersections was performed by the convolution of the image with a kernel (3D cross with r=3 mm and bars width 1.5 mm [6]). Convolution images were thresholded. A regular grid of control points was then placed, and residual tilts were removed [7]. We could then create the distortion field by calculating vectors between control and measured points. In case of missing measured points, interpolation was applied [8].

3. Results

More than 95 % of 2002 vertex were detected in all acquisitions. Fig. 1 shows the distortion field as a 3D heat map that includes 3 planes where the isocenter is visible. Graphs show errors of phantom points in mm as a function of the radial distance to the iso-center. Scattered points, in blue and red, correspond to,  respectively, the center and superior laser positioning. While the laser marking was on the phantom center, mean error was 0.55 mm (without filter) and 0.42 mm (with filter) for FSPGR, 0.59 mm and 0.41 mm for MP2RAGE. For the superior laser marking, mean error was 0.79 mm and 0.59 mm for FSPGR, 1.48 mm and 0.68 mm for MP2RAGE.

4. Conclusion

Positioning of the object to be imaged is critical in terms of geometric distortions. Centering in the tunnel should definitely be done with respect to the region of the interest when the aim is to accurately target in stereotaxy procedures. Moreover, constructors’ distortions-correction filters helps efficiently to reduce the geometric distortions due to the gradient non-linearities and should be systematically activated.


This study was partially supported by General Electric.


[1] K Wachowicz et al., Medical Physics, 39.5, 2659-2668, 2012

[2] Weygand et al. , International Journal of Radiation Oncology,95.4 1304-1316, 2016

[3] Baldwin and Lesley N. et al. Medical Physics,34.2, 388-399, 2007

[4] Fedorov, Andriy et al. Magnetic resonance imaging,30.9,1323–1341, 2012

[5] Mirebeau, Jean-Marie et al. arXiv: 1503.00992., 2015

[6] Huang, Ke et al. Physics in Medicine and Biology, 61.2, 774-790, 2016

[7] Besk, McKay et al. IEEE Trans. Pattern Anal. Mach. Intell, 14.2, 239–256,1992

[8] Shepard. ACM ’68 New York, 517-524, 1968

16:05 - 16:15 #16333 - O104 Single-subject connectomics with morphometric similarity networks in patients with Parkinson’s disease prior to deep brain stimulation.
O104 Single-subject connectomics with morphometric similarity networks in patients with Parkinson’s disease prior to deep brain stimulation.



Over the past decade there has been a universal movement within the neuroscience community seeking to describe the brain’s wiring diagram or ‘connectome’. Numerous advances have been made through this endeavour including the understanding that broad rules govern network function over a range of scale, species, and phenotypes. Application to clinical practice however has been made difficult by the requirement for specialist neuroimaging data and the lack of methods for single subject analysis. Recently, Morphometric Similarity Networks (MSN) have been proposed, based upon shared variance in structural features of the grey matter, as a proxy for network connectivity at the single subject level. This study seeks to validate the use of MSNs in clinical practice for the first time in a cohort of patients with Parkinson’s disease prior to deep brain stimulation. The aim was to derive the core connectomic features of these MSNs and quantitatively validate them against universal network features described in the literature.




A retrospective cohort study was performed of a consecutive series of patients with Parkinson’s disease due to undergo deep brain stimulation of either the GPi or STN. All patients had 3 Tesla MRI scans with MPRAGE sequences. Cortical surface reconstructions were performed with Freesurfer (v6.0) and parcellated with the Desikan-Killiany 68 node template. Morphometric Similarity Networks were constructed based upon surface area, grey matter thickness, mean curvature, Gaussian curvature, intrinsic curvature, and folding index. Connectome construction was based on statistical dependencies (e.g. Pearson correlation) between network nodes with bootstrap-based thresholding using the false discovery rate. General network features were computed including degree distribution fit and small worldness. Weighted network measures were computed including consensus hubs, optimised modularity, and node versatility. 




In total 28 participants (17 STN, 11 GPi) met the inclusion criteria (12 female). Mean age was 63 years, mean disease duration was 12 years, and mean follow-up was 11 months. Network features were generally robust to construction methods including methods of statistical dependency and thresholding. Small world features were present throughout (Humphries ~2.0, Latora ~1.8, Telesford 0.2) indicating numerous closely connected local communities interconnected by short-cuts between them, thereby balancing features of both lattice-based and random graphs. A variety of hubs were identified predominantly in higher association cortices, consistent with their role in global network integration and higher cognitive function. Consensus modularity and node versatility were used to define an optimal community partition of the network into approximately 5-10 modules. 




Single subject structural connectomics in patients with Parkinson’s disease awaiting deep brain stimulation is practical using clinically acquired data and offers realistic network representations or connectomes. Analysis of Morphometric Similarity Networks identifies the quintessential network features of small worldness, higher association area hubs, and distinct community partitions. These described network features are consistent with those described in other networks and connectome over a range of imaging modalities. Additionally, they are robust to variation in the network analysis construction. With these models there is the potential to describe and ultimately predict a range of clinical features not hitherto accessible including for example neuropsychiatric symptoms and higher cognitive function. 


Michael HART (Cambridge, UK), Rafael ROMERO-GARCIA, Philip BUTTERY, Robert MORRIS, Jakob SEIDLITZ
16:15 - 16:25 #16401 - O105 Tractography reconstruction of hyperdirect pathway fibers connecting the primary motor cortex to the subthalamic nucleus.
O105 Tractography reconstruction of hyperdirect pathway fibers connecting the primary motor cortex to the subthalamic nucleus.

Introduction Deep Brain Stimulation (DBS) of the subthalamic nucleus (STN) is an established therapy to relieve motor symptoms in advanced Parkinson’s disease patients. The treatment is likely to affect white matter fibers of the hyperdirect pathway that convey excitatory effects from cortical motor areas to the STN. Recent studies using single tensor deterministic tractography and multi-fiber probabilistic tractography have demonstrated the feasibility of identifying the hyperdirect pathway in diffusion MRI (dMRI) data (1-4). Through the Human Connectome Project (HCP), multi-shell dMRI data have been made available for mapping white matter anatomy at an unprecedented resolution (5). We propose to investigate the use of multi-fiber deterministic tractography in high-resolution HCP data to reconstruct the trajectory of hyperdirect pathway fibers connecting the primary motor cortex to the STN.

Methods We used T1, T2 and dMRI datasets acquired on five HCP subjects (5). The dMRI scans were acquired using single-shot 2D spin-echo multiband EPI sequence using 90 gradient directions, 3 b-values (1,000 s/mm2, 2,000 s/mm2, 3,000 s/mm2), 1.25 mm slice thickness and 1.25 image resolution. For each subject, we used the 3D Slicer open-source software ( to define regions of interest (ROIs) in the primary motor cortex, internal capsule and STN on the T1 images for tractography purpose. We used an automated atlas-based segmentation approach based on the Yeb Atlas to generate 3D meshes of the STN (6-7), and we post-processed the 3D meshes to integrate anatomical information contained in the dMRI data. A tractography workflow using a multi-compartment model of diffusion and multi-fiber deterministic tractography was used to reconstruct the tracts (8). Fiber tracking was performed with 20 seeds/voxel, step size of 0.5 mm and curvature threshold of 55 degrees. The anatomical accuracy of the reconstructed tracts was evaluated by four neuroanatomical experts using the DTI challenge methodology (

Results Hyperdirect pathway fibers connecting the primary motor cortex to the STN were successfully reconstructed in all five subjects (Fig.1). The evaluation of the tracts by neuroanatomical experts demonstrated that the tractography results were in agreement with the expected anatomy.

Conclusion We have demonstrated that multi-fiber deterministic tractography combined with high-resolution multi-shell dMRI data can be used to reconstruct the three-dimensional trajectory of hyperdirect pathway fibers in individual subjects. Advanced tractography techniques combined with cutting-edge dMRI data have the potential to provide novel clinical research tools for personalized white matter mapping in the DBS treatment of Parkinson’s disease.



1.Chen Y et al. The role of the cortico-subthalamic hyperdirect pathway in Deep Brain Stimulation for the treatment of Parkinson's Disease: A diffusion tensor imaging study. World Neurosurgery 2018. 2.Avecillas-Chasin JM et al. Tractographical model of the cortico-basal ganglia and corticothalamic connections: Improving our understanding of Deep Brain Stimulation. Clin Anat 2016;29(4):481-92.

3.Lambert C et al. Confirmation of functional zones within the human subthalamic nucleus: patterns of connectivity and sub-parcellation using diffusion-weighted imaging. Neuroimage. 2012;60:83-94.

4.Akram H et al. Subthalamic deep brain stimulation sweet spots and hyperdirect cortical connectivity in Parkinson's disease. Neuroimage 2017;158:332-345.

5.David et al. The WU-Minn Human Connectome Project: An overview. NeuroImage 2013;80:62-79.

6.D'Albis T et al. PyDBS: an automated image processing workflow for DBS surgery. Int J Comput Assist Radiol Surg 2015;10(2):117-28.

7.Bardinet E et al. A 3D histological atlas of the human basal ganglia. II. Atlas deformation strategy and evaluation in deep brain stimulation for Parkinson disease. J Neurosurgery 2009;110(2), 208–19.

8.Pujol S et al. In vivo Exploration of the connectivity between the subthalamic nucleus and the globus Pallidus in the human Brain using multi-fiber tractography. Front Neuroanat 2017;19;10:119.

Figure. 1 Multi-fiber tractography reconstruction of hyperdirect pathway fibers. The figure shows the trajectory of white matter fibers connecting the primary motor cortex to the STN. An axial T1 image with 3D surface models of the STN (light orange), GPi (dark orange) and GPe (green) are displayed for anatomical reference.

16:20 - 16:30 #16159 - O106 Deep brain stimulation: Patient-specific electrical field simulation and tractography.
O106 Deep brain stimulation: Patient-specific electrical field simulation and tractography.


Electrical field (EF) simulation is used to evaluate the volume of tissue activated (VTA) for patients undergoing deep brain stimulation (DBS). The aim is to develop a workflow for patient-specific EF simulation which integrates tractography reconstruction of the dentato-rubro-thalamic tract (DRT).


Preoperative diffusion MRI (dMRI) (b=800s/mm2, 32 directions, voxel size=1.75x1.75x2 mm) and stereotactic T2-weighted scans (TR=8000ms, TE=80 ms, voxel size=0.5x0.5x2 mm) using Leksell stereotactic system (Elekta instrument AB, Sweden) were acquired on a patient with essential tremor (ET). The study was approved by the local ethics committee (2012/434-31).

The dMRI data were corrected for eddy current distortion and diffusion models were created with bedpostx (fibres/voxel=2). Probabilistic tractography reconstruction of the DRT was generated using probtrackx2 (FSL v.5.0, FMRIB Analysis Group, University of Oxford, UK). Tracts were seeded in the precentral gyrus (number of samples=5000, curvature threshold=0.2), and the superior cerebellar peduncle and dentate nucleus were used as waypoint masks. Projections to contralateral cerebrum or ipsilateral cerebellum were excluded.

An in-house software ELMA (Ver. 2.4, Dept. of Biomedical Eng., Linköping University, Sweden) was used for intensity-based tissue segmentation of the preoperative T2 images, with a volume of interest of 100x100x42 mm3. The tissues were assigned tabulated conductivity values: grey matter (0.123 S/m), white matter (0.0754 S/m), CSF (2.0 S/m) and blood (0.7 S/m)[1]. SurgiPlan (Elekta instrument AB, Sweden) was used to co-register the preoperative MRI with postoperative CT. From the co-registered images, Leksell coordinates from the artefacts were retrieved and used for positioning the DBS-leads during simulation. A model of the DBS lead was created (6172, Abbott Lab., Il., USA) [2]. The EF was simulated, using the finite element method (Comsol Multiphysics Ver. 5.2, COMSOL AB, Sweden). The workflow was exemplified with the ET patient operated in Zona incerta (Zi). Simulations were performed in ring mode (L: contact 2, R: contact 10, Fig. 1a) and by using each segment of the contacts separately (n = 8). Stimulation parameters were set to: pulse width=60 µs, frequency=140 Hz, amplitude L=1.3mA and R=1.0mA. The rotation of the electrodes within the patient was not known meaning the evaluation from different contact segment is only theoretical. An isolevel of 0.2 V/m, corresponding to axon diameters of approximately 3 µm and larger [3], was used to visualize the VTA together with the generated DRT. The VTA and its expansion was evaluated for all simulations as well as the common volume with the generated DRT.


A workflow for patient-specific EF simulation using tractography reconstruction of the DRT was developed. The generated DRT tracts were passing through Zi and crossed in the lower part of the midbrain (Fig. 1b). The size of the VTA for left side (ring mode and each segment) varied between 55-62 mm3 and for right side 34-37mm3. One segment of the contact will steer the EF and expand VTA further in that direction compared to ring mode. The variation in the common volume of VTA and DRT was higher than the variation of the VTA (L: 18-33 mm3, R: 6-21 mm3). For both sides the highest common volume of VTA and DRT was retrieved using one of the segments.


A first version of a workflow for combined patient-specific EF simulations and tractography reconstructions has been developed. Further work will focus on improvement of the dMRI protocol and refinement of the patient-specific simulation method. As the DRT is being considered as a potential DBS target, this study indicates that the steering electrode could be beneficial in activating as many of the fibres within the tract as possible.


[1] M. Åström, L.U. Zrinzo, S. Tisch, E. Tripoliti, M.I. Hariz, K. Wårdell, Method for patient-specific finite element modeling and simulation of deep brain stimulation, Medical & Biological Engineering & Computing 47(1) (2009) 21-28.

[2] F. Alonso, M.A. Latorre, N. Göransson, P. Zsigmond, K. Wårdell, Investigation into Deep Brain Stimulation Lead Designs: A Patient-Specific Simulation Study, Brain Sciences 6(3) (2016) 39.

[3] M. Åström, E. Diczfalusy, H. Martens, K. Wårdell, Relationship between Neural Activation and Electric Field Distribution during Deep Brain Stimulation, IEEE Transactions on Biomedical Engineering 62(2) (2015) 664-672.

Teresa NORDIN (Linköping, SWEDEN), Peter ZSIGMOND, Sonia PUJOL, Carl-Fredrik WESTIN, Karin WÅRDELL
16:30 - 16:35 #16234 - O107 Infusion-enhanced, MRI-guided surgery for deep brain stimulation.
O107 Infusion-enhanced, MRI-guided surgery for deep brain stimulation.

Background.  Image-guided surgery is becoming increasingly utilized in neurosurgery with the availability of intraoperative CT and MRI.  Now neurosurgeons are relying on image-guided technologies for laser interstitial thermal therapy (LITT), focused ultrasound (FUS), deep brain stimulation (DBS) and brain tumor resections.  Since initial clinical observations in patients with edema, we have been developing a technique to enhance the MRI visualization of deep brain structures with convection-enhanced delivery (CED) of CSF.  Preclinical testing in the laboratory suggests feasibility and safety; and has provided the rationale for an FDA-approved clinical trial in eight patients with Parkinson disease (PD).    

Methods.  MRI-guided DBS was planned for patients with PD under general anesthesia.  Target coordinates were first planned based on traditional methods using direct visualization and a stereotactic atlas.  Infusion of 500 microliters of the patients own CSF, obtained from a lumbar puncture, was then performed at 5 microliters / min while obtaining serial MR images.  Following infusion, the target was reassessed based on direct visualization of the target.

Results.  In our first patient, CED of 500 microliters of CSF into the globus pallidus was safe and highlighted the border of the internal nucleus.  The target was adjusted 0.6mm posterior, 1.0mm lateral, and 0.9mm superior - altering the vector in positioning by 1.5mm. 

Conclusion.  The convective infusion of autologous CSF seems safe and improves MRI visualization during deep brain stimulation (DBS) electrode placement.  This could supplant the need for inference from published atlases and potentially improve image-guided DBS surgical outcomes.

Aaron BOND, Tony WANG, W. Jeff ELIAS (Charlottesville, USA)
16:35 - 16:40 #16383 - O108 Probabilistic mapping of the cortical connections of the active contact regions in Vim deep brain stimulation (continuation of the previous study).
O108 Probabilistic mapping of the cortical connections of the active contact regions in Vim deep brain stimulation (continuation of the previous study).

Objective: The ventralis intermedius nucleus (Vim) of the thalamus has been known as a notorious target of stereotactic surgery. At the present, the mechanism of action and connectivity of the Vim and the ventralis oralis posterior (Vop) are still unidentified. The aims of this continuation of the previous study are to collect more information about active contact regions in successfully treated cases. Furthermore, with the calculated and averaged cortical connectivity maps, we probably able to create more usable individual tractography for Vim-DBS targeting. 

Methods: In this recent study all of our 18 patients have suffered from Essential tremor and they have been treated with Vim-DBS (19 operations, 30 leads). In this group, the tremor amplitude reduction reaches 90%. The position of the active contact(s) have been allocated by preoperative MR - postoperative CT fusion, after registration in the MNI152 standard-space. The number of connections of the cortical voxels to the active contact region were calculated by MRIB Software Library with 5000 cycle number.

Results: In our investigation the active contacts are located mainly below the AC-PC plane (MD –1,33mm, SD±1,68). It seems that the active contacts draw a pathway instead of a spherical target volume. The probabilistic classification’s determined motor and premotor thalamus did not contain the Roi of the active contacts. With the utilization of the Harvard-Oxford cortical structural atlas, the frontal projections of the Roi did not correspond to the motor and premotor connections of the thalamus. The connectivity to the supplementary motor cortex and to the medial part of the superior frontal cortex seemed to play more significant role for the efficient stimulation than we previously thought.

Conclusions: In our investigation the drawn tractography is consistent with the published anatomical studies, for example the dentato-rubro-thalamo-cortical pathway, but the pattern of cortical connection reveals that the non-invasive Diffusion tensor imaging (DTI) based thalamus segmentation still needs some modifications to help us for Vim-DBS targeting.

16:40 - 16:55 Image-verified DBS as The Standard of Care. Ludvic ZRINZO (London, UK)
16:55 - 17:00 Discussion.

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
15:00 - 17:00

Parallel Session 12
Movement Disorder

Moderators: Stephan CHABARDÈS (head of the unit) (GRENOBLE, FRANCE), Claudio POLLO (Deputy Chief Doctor) (Bern, SWITZERLAND), Erlick PEREIRA (Consultant Neurosurgeon) (London, UK), Jordi RUMIA (Coordinator. Adult and Paediatric Functional Neurosurgery Program) (Barcelona, SPAIN)
15:00 - 15:15 Directional electrodes. Stephan CHABARDÈS (head of the unit) (GRENOBLE, FRANCE)
15:15 - 15:20 #14911 - O109 Changes of functional and dysfunctional impulsivities after deep brain stimulation of the subthalamic nucleus : predictive factors and electrodes location.
Changes of functional and dysfunctional impulsivities after deep brain stimulation of the subthalamic nucleus : predictive factors and electrodes location.

Background: Although deep brain stimulation (DBS) of the subthalamic nucleus (STN) is routinely proposed to treat Parkinson’s disease (PD) patients with motor fluctuations, its influence on postoperative global impulsivity or impulse control disorders (ICD) is controversial. We studied changes of impulsivity in PD patients treated by DBS-STN, by considering the two dimensions of impulsivity defined by Dickman, namely functional (“physiological”) impulsivity (FI) and dysfunctional (“pathological”) impulsivity (DI).

Methods: We studied retrospectively pre and postoperative data of 33 idiopathic PD patients with levodopa related-motor complications treated by DBS-STN, including UPDRS score, treatment (including dopaminergic agonists intake), cognitive functions (FAB and Mattis scales, apathy inventory), mood (MADRS) and occurrence of ICDs. Impulsivity was assessed using the Dickman scale, who distinguishes two dimensions: the FI that reflects the ability to react quickly when the situation requires it, and the DI who reflects the absence of preliminary reflection in the action, even when the situation requires it. Localization of the stimulating contacts was studied using a deformable histological basal ganglia atlas merged on postoperative MRI. 

Results: Six months after surgery, we observed a significant increase (p<0,001)of DI (mean pre- and post-operative DI scores 9 +/-1,6 and 3,5 +/-2,4) but no change in FI (mean pre- and post-operative FI scores 6,2 +/-2,7 et 5,8 +/-2,6). Factors associated with DI score’s increase >2 were: low preoperative FAB score (p=0,03), high preoperative UPDRS III score (p=0,02) and location of the left electrode in the ventral STN (p=0,012). 

 Conclusion: Our study suggests that DBS-STN might influence differently the two dimensions of impulsivity, by worsening the pathological impulsivity without modifying the physiological impulsivity. Severity of the PD, preoperative impairment of frontal functions and ventral location of the electrode favored worsening of dysfunctional impulsivity. Patients with these contributing factors should beneficiate from a closer psychiatric monitoring. 

Robin KARDOUS, Bruno GIORDANA, Caroline GIORDANA, Michel BORG, Jean Jacques LEMAIRE, Denys FONTAINE (Nice)
15:20 - 15:25 #14952 - O110 The influence of brain atrophy on targeting accuracy and outcome in robot-assisted, guide tube-directed DBS for Parkinson’s Disease.
O110 The influence of brain atrophy on targeting accuracy and outcome in robot-assisted, guide tube-directed DBS for Parkinson’s Disease.



Traditionally brain atrophy has been considered a relative contraindication to DBS surgery due to an increased risk of CSF egress, brain shift, targeting error and surgical complications. The aim of this study was to determine the effect of brain atrophy on targeting accuracy and outcome in a cohort of PD patients undergoing DBS surgery using a highly accurate robot-assisted and guide-tube directed method designed to minimised CSF loss and brain shift.




All surgical procedures were performed using a robot-assisted guide tube-directed method (Renishaw PLC, Gloucs., UK) with on-table fluoroscopic scanning for image-guided targeting verification (O-arm, Medtronic, MN, USA). The drilling process utilised a 1.8mm diameter drill which was intended to minimise the risk of CSF egress and subsequent brain shift. Guide tubes were inserted using a track dilation technique in order to create a pre-formed track. Brain atrophy was quantified by determining the brain parenchymal fraction (BPF) in 113 consecutive PD patients using the MICO method for segmentation of grey matter, white matter and CSF, with a low BPF reflecting increased brain atrophy. Targeting accuracy was determined by comparing the planned guide tube and stylet position with intra-operative fluoroscopic scans following implantation. Vector, radial and depth targeting errors were determined and correlated with BPF using Matlab software (Mathworks, Cambridge, UK) and Spearman's rank correlation analysis. Clinical outcome data was prospectively collected and BPF correlated with UPDRSIII scores at 6 months following surgery. Brain shift was also quantified in a subgroup of 73 patients by comparing the pre-operative and intra-operative position of blood vessels, and correlated with the degree of brain atrophy.




Using the MICO method for brain segmentation, the BPF of PD patients within this cohort ranged from 0.53 to 0.89 (mean 0.78). There was a weakly significant correlation between increased brain atrophy and both radial and vector targeting error, with a trend towards greater error in patients with BPF<0.7. However, the maximum vector and radial targeting error was less than 1.6mm in all patients, with a mean error of 0.78mm. There was no correlation between the degree of brain atrophy and percentage change in UPDRSIII scores at 6 months. No correlation could be found between the degree of brain atrophy and intra-operative brain shift (r=0.18). Two patients suffered brain haemorrhages within this cohort – the BPF in one patient was high (0.82) and lower in the second patient (0.67).       




Using a robot-assisted and guide tube-directed surgical technique designed to minimise CSF egress, targeting remained highly accurate in patients with low BPF and increased brain atrophy, with no correlation found between the degree of atrophy and clinical outcome at 6 months post-operatively. Furthermore, the degree of brain atrophy could not be correlated with intra-operative brain shift in this cohort. Further studies are required to determine whether patients with BPF<0.7 should be counselled for increased risk of targeting error, or whether a “threshold” BPF can be identified which may be used as a selection criterion for patients considered for DBS.  Comparative studies of the effect of brain atrophy on targeting accuracy and outcome when DBS surgery is performed using conventional burr holes with the robot-assisted guide tube-directed method described here are also warranted.

Neil BARUA (Bristol, UK), Mariusz PIETRZYK, Catherine MORAN, Max WOOLLEY, Steven GILL
15:25 - 15:30 #15046 - O111 Bilateral Deep Brain Stimulation in the caudal Zona incerta for Parkinson's disease - 1-year evaluation.
O111 Bilateral Deep Brain Stimulation in the caudal Zona incerta for Parkinson's disease - 1-year evaluation.

Background: Deep brain stimulation (DBS) is an established treatment for advanced Parkinson’s disease (PD). The most commonly used targets are the subthalamic nucleus, the pallidum and the ventrolateral thalamus. The area below the thalamus, known as the posterior subthalamic area (PSA)/caudal Zona incerta (cZi) has been shown to alleviate PD symptoms such as rigidity and tremor, when subjected to ablation or high-frequency stimulation. The published experience is, however, limited and consists to a large part of unilateral procedures in patients with tremor dominant symptoms.

Aim: To evaluate the effect of bilateral cZi-DBS in a group of patients with PD fulfilling the criteria for bilateral STN-DBS.

Method: The patients in this study were previously included in a randomised blinded trial of cZi DBS versus best medical treatment, with a 6 months follow-up (Blomstedt P, Stenmark Persson R, Hariz G, et al. Deep brain stimulation in the caudal zona incerta versus best medical treatment in patients with Parkinson’s disease: a randomised blinded evaluation. JNNP 31 January, 2018; Epub ahead of print). In the present study we analysed the results of surgery for the whole cohort at 6 and 12 month after surgery. Main outcomes were differences in motor symptoms (Unified Parkinson’s Disease Rating Scale, UPDRS-III) and in quality of life (Parkinson’s Disease Questionnaire , PDQ-39) between baseline and follow-ups. The patients were evaluated on/off medication at baseline before surgery and on/off medication, on/off stimulation at 6 and 12 months. Preoperative T2-weighted images was used for intensity-based image segmentation and conductivity assignment (ELMA Version 2.4, Department of biomedical engineering, Linköping University, Sweden) based on tabulated conductivity values resulting in a brain model. Patient-specific electrical field simulation of the brain model will be performed using the finite element method (Comsol Multiphysics Ver 5.2, COMSOL AB, Stockholm, Sweden). Friedman’s test was used for statistical evaluation of non-parametric values and the Wilcoxon signed rank test as a post-hoc analysis. Analysis of variance for repeated measurements was used for continuous variables with the Bonferroni correction method as a post hoc test. A p-value ≤0.05 was considered statistically significant.

Fifteen patients (4 females) were included. The mean age at surgery was 58.6±2.5 years. Levodopa-equivalent daily dosage (LEDD) were 1248±685 mg at baseline.  The presented results are means ± standard deviation in off-medication state, unless stated otherwise.  
UPDRS-III scores improved from 37.5±11.2 at baseline to 21.2±6.6 (44%) and 22.7±7.5 (40%) on stimulation at 6 and 12 months, respectively (p≤0.0005). Scores of on versus off stimulation showed  an improvement of 49% at 6 months and 47% at 12 months (p≤0.0001).  
Tremor score improved from 9.4±5.4 at baseline to 0.8±1.1 (92%) and 1.2±1.3 (87%) on stim, at 6 and 12 months after surgery (p≤0.0001). Rigidity scores improved from 7.6±2.7 to 5.4±2.7 (29%) and to 5.5±2.5 (28%) (p≤0.005). Akinesia scores improved from 13.0±5.0 to 9.0±4.3 (31%) at 6 months and 9.9±4.8 (24%) at 12 months (p≤0.01). Axial scores improved by 19% at 6 months (p≤0.02) but were no longer significantly changed at 12 months. For PDQ-39, ADL was improved from 28.1±22.2 to 20.8±19.1 (26%, p≤0.05). Stigma improved from 23.4±19.3 to 11.3±15.9 (52%, p≤0.01).

There were no significant changes concerning dyskinesia scores or LEDD. Mean stimulation parameters were 2.3±0.5 V, 66±12.2 µs and 148±13.4 Hz at 6 months and did not change over time. In relation to the midcommissural point, the active contacts were 11.8±1.2 mm lateral, 6.4±1.4 mm posterior and 1.9±1.7 mm inferior.

Discussion: The cZi has been suggested as an alternative target for DBS for tremor, including for PD. In this study there was robust effect at one year on tremor, and moderate on akinesia, as well as improvement in ADL and stigma. At 12 months cZi DBS did not affect axial symptoms, dyskinesia scores or LEDD. The cZi as a target for DBS in PD is an addition to the established targets, allowing perhaps to tailor the surgery to the needs of the individual patients. Further analysis in the project will include evaluation of activation volume in reference to the anatomy. Further longer term studies are needed to determine the role of cZi DBS in the surgical therapeutic armamentarium of PD.

Rasmus STENMARK P. (Umeå, SWEDEN), Teresa NORDIN, Gun-Marie HARIZ, Patric BLOMSTEDT
15:30 - 15:35 #16134 - O112 Fiber-tracts associated with the alleviation of bradykinesia in deep brain stimulation for Parkinson´s disease.
O112 Fiber-tracts associated with the alleviation of bradykinesia in deep brain stimulation for Parkinson´s disease.


Deep brain stimulation becomes increasingly important in treating the motor symptoms of Parkinson’s disease. To improve target planning in deep brain stimulation, we investigated the connectivity between the stimulated volume and cerebral white matter tracts to find out, which tracts have a positive effect on bradykinesia in Parkinson’s disease.


This retrospective study was based on 21 patients with Parkinson’s disease who received bilateral deep brain stimulation in the subthalamic nucleus. 18 of these patients initially presented with bradykinesia symptoms existent in 31 hemibodies. Since each lead electrode has four contacts, 124 lead contacts in the associated contralateral brain hemispheres were investigated. The effectiveness in reducing bradykinesia was individually assessed for every lead contact. Furthermore, we performed probabilistic tractography to visualize cerebral fiber-tracts showing connectivity to the stimulated brain volume of each individual electrode contact. Our calculations were carried out using preoperatively acquired diffusion weighted magnetic resonance imaging (64 gradient directions, 12-channel head-coil).


Out of the 124 contacts, 92 (74.2%) showed a reduction of the contralateral bradykinesia symptoms of 50% or more and were therefore considered clinically effective. Clinically effective and ineffective contacts were compared for their connectivity patterns using the computed fiber-tracts. The statistical analysis showed that the effective contacts were significantly (p < 0.05) more often associated with the ipsilateral superior cerebellar peduncle and the ipsilateral dentate nucleus, which partly define the course of the ipsilateral, non-crossing part of the cerebello-thalamo-cortical pathway. Furthermore, the anterolateral fasciculus, the medial forebrain bundle, the zona incerta, the pedunculopontine nucleus and the medial part of the internal capsule’s anterior limb were significantly positively associated. Significant negative correlations existed with the contralateral medial cerebellar peduncle, the central pons and the motor cortex representing the descending cortico-ponto-cerebellar pathway.


A connectivity-based approach may improve target planning in deep brain stimulation. In the case of bradykinesia in Parkinson’s disease, the cortico-cerebellar loop seems to play a key role. Our results showed that a connection to its ascending portion, the ipsilateral cerebello-thalamo-cortical pathway, strongly correlated with the alleviation of bradykinesia, while stimulation of the descending cortico-ponto-cerebellar pathway had no positive influence.

Quirin STROTZER (Regensburg, GERMANY), Judith ANTHOFER, Rupert FALTERMEIER, Alexander BRAWANSKI, Claudia FELLNER, Anton BEER, Juergen SCHLAIER
15:40 - 15:45 #16194 - O114 Are there advantages of dbs on the awake pd patient for the clinical outcome?
O114 Are there advantages of dbs on the awake pd patient for the clinical outcome?


The need of using multiple trajectories and intraoperative testing in an awake patient during DBS for Parkinson´s disease is discussed controversially. In our department it was done in all cases. We aimed to find out if there is a benefit of such procedure in the clinical outcome.


We retrospectively analysed all data of patients, who have undergone DBS-surgery for Parkinson`s disease in our department in 2012 and 2014. We recorded the number of microelectrodes, the trajectory of the permanent electrode, the reasons for avoiding the central trajectory and the clinical outcome (UPDRS Part III). The intraoperative effect was defined as the improvement of the rigor estimated in % by a neurologist. The clinical outcome was measured as the difference of the preoperative UPDRS III under L-dopa medication and the postoperative UPDRS III under L-Dopa medication and stimulation. Possible influencing factors on the clinical outcome were estimated with a linear regression model (p<0,05).


In the study period 67 patients were treated with bilateral DBS of the STN because of Parkinson`s disease (134 implantations of permanent electrodes). In 92,88% of the 134 implantations 3 or more trajectories were used for microrecording and intraoperative testing. For implantation of the permanent electrode the central trajectory was used in 58%. In the remaining patients (42%), we used in 56,36% the anterior, in 1,81% the medial and in 41,81% the lateral trajectory. The reason for not using the central trajectory was a better intraoperative effect in 65,38%. Changing the trajectory due to a better intraoperative effect was found to be an independent, significant (p= 0,001) predictor for the clinical outcome with the coefficient factor 8,32 on the right and 9,93 on the left side.


Using multiple trajectories lead to a deviation from the central trajectory in 42 % of the implantations, predominantly due to a better intraoperative effect. This intraoperative testing has turned out to be significantly positively correlated with the clinical outcome (even after controlling for the other variables). This result is in line with the hypothesis that deviating from the central trajectory due to intraoperative testing leads to a better clinical outcome. However, a matched pair analysis between the operation with awake and intubated patients, which would allow to test for a causal relationship, has not been done yet but belongs to our agenda.

Soeren K. HAUCK (Hannover, GERMANY), Steffen PASCHEN, Jan VOGLER, Michael SYNOWITZ, H. Maximilian MEHDORN, Daniela FALK
15:45 - 15:50 #16198 - O115 Bilateral thalamic deep brain stimulation for post-encephalitic movement disorders: A case report.
O115 Bilateral thalamic deep brain stimulation for post-encephalitic movement disorders: A case report.

Introduction: Deep brain stimulation (DBS) is an established treatment for essential tremor, Parkinson’s disease and dystonia. DBS is also effective in several symptomatic movement disorders, yet, its therapeutic effect on post-encephalitic symptoms has been under-reported.

A case description: A 47–year-old man presented with progressive severe axial and limb tremors for 6 months after resolution of his autoimmune encephalitis. On examination, he showed resting, intentional, and task-related tremors in his trunk and bilateral limbs. He also had dysarthria and moderate torticollis to the right side. His preoperative TETRAS scale score was 44. As the tremors being medically refractory, he underwent bilateral ventral intermediate nucleus (Vim) DBS surgery. The coordinates for Vim were 13.5mm lateral and 5.8mm posterior to the midcommissural point on the anterior and posterior commissural plane. Intraoperative microelectrode recording showed tremor cells from 8.4mm to 2.0mm above the target. The DBS electrodes were implanted to the targets to stimulate these recorded points. After an initial DBS programming, postoperative TETRAS scale score improved to 30.5 by ameliorating his tremor. His dysarthria persisted, yet moderate torticollis disappeared.

Discussion: Bilateral Vim DBS can be an effective treatment for severe post-encephalitic movement disorders.

Yusuke NAKAJIMA (Seidocho, Nakagyo-ku, Kyoto-shi, Kyoto, JAPAN), Namiko NISHIDA, Shigeto NAGAO, Akihiko OZAKI, Koichi IWASAKI, Hiroki TODA
15:50 - 15:55 #16199 - O116 Endoscopic third ventriculostomy for parkinsonism due to hydrocephalus caused by aqueductal stenosis after midbrain hemorrhage: A case report.
O116 Endoscopic third ventriculostomy for parkinsonism due to hydrocephalus caused by aqueductal stenosis after midbrain hemorrhage: A case report.

Introduction: Movement disorders related to hydrocephalus are frequently expressed as Parkinsonism, yet anatomical substrates are fairly unclear. We report here a case of post- hemorrhagic aqueductal stenosis and discuss about the origins of symptoms.  

A case description: A 23–year-old man presented with progressive bradykinesia, ataxia and gait disturbance 1.5 months after acute-phase ventricular drainage treatment for left midbrain hemorrhage ruptured into third ventricle. On examination, he showed masked face, vertical gaze palsy and resting, intentional, and task-related 3-4 Hz tremors in his right upper limb. He also had dysarthria and pyramidal tract sign of right side. His preoperative MDS_UPDRS 3 score was 68. MRI showed triventricular hydrocephalus due to edematous obstruction of aqueduct surrounding ruptured hematoma cavity, which was confirmed endoscopically. Dopamine transporter scan showed decreased uptake in left putamen. After third ventriculostomy, both ventriculomegaly and midbrain edema subsided. His symptoms were alleviated, and MDS_UPDRS 3 score improved to 19 at discharge. Dopaminergic agent was precluded due to its adverse effect. He returned to the work within a year, with residual minimum gaze palsy.

Discussion: Hydrocephalic Parkinsonism is composed of complex symptoms according to affected white matter tracts. In this case, pyramidal tract, medial longitudinal fasciculus, dentate-rubro-thalamic tract and nigrostriatal tract were possibly affected by edema, and their dysfunctions were partially reversible.

Namiko NISHIDA (Osaka, JAPAN), Yasunori NAGAI, Naoya YOSHIMOTO, Hirokuni HASHIKATA, Masanori GOTO, Kenichi KOMATSU, Hidemoto SAIKI, Hiroki TODA, Sadayuki MATSUMOTO, Koichi IWASAKI
15:55 - 16:00 #16203 - O117 Correlation between the volume of activated tissue, its overlap with the pyramidal tract, and the intra-operative side effect threshold.
O117 Correlation between the volume of activated tissue, its overlap with the pyramidal tract, and the intra-operative side effect threshold.

Background: The volume of activated tissue (VAT) virtualize the current provide to the DBS lead on brain structures. The objective of the present study was to assess the correlations between the VAT, activation of the corticospinal tract, and the intra-operative side effect (ISE) threshold.

Methods: This double-blind, single-center study was performed between September, 2016, and July, 2017. We identified two groups for statistical analysis: the entire study population, and a subset of patients with additional diffusion tensor imaging (DTI) data for determining the location of the pyramidal tract. We determined the intensity threshold at which the VAT reached the border of the target nucleus (referred to as VATn) and the intensity threshold when the VAT reached the pyramidal tract (referred to as VATndti). In each group of patients, we studied the correlation between the ISE threshold and the VATn or VATndti threshold.

Results: Fifteen patients were included in the study. In both groups, there was a significant correlation between the VAT intensity threshold and the ISE threshold (p=0.018; r=0.31 for VATn in the entire study population). In the subset of patients with valid tractography data, the correlation was stronger (p=0.002; r=0.5 for VATndti).

Conclusion: The present study is the first to show a relationship between the intensity threshold as determined by the use of the VAT and the intra-operative side effect threshold. The correlation between the clinical features and the VAT appear stronger when the model was based on a combination of high-resolution anatomic data and interpretable DTI data.


Katia BUNAUX (Amiens), Mélissa TIR, Jean-Marc CONSTANS, Jean-Michel MACRON, Pierre KRYSTKOWIAK, Michel LEFRANC
16:00 - 16:05 #16219 - O118 Pallidotomy and pallidal deep brain stimulation in 2 patients with post-hyperglycaemic chorea-ballism.
O118 Pallidotomy and pallidal deep brain stimulation in 2 patients with post-hyperglycaemic chorea-ballism.


Hyperglycaemia has recently gained attention as the second most common cause of chorea-ballism. Although usually a self-restricted movement disorder resolving with glycemic correction, persistent chorea-ballism can be difficult to manage medically.


Case report

We report on 2 patients with post-hyperglycaemic chorea-ballism treated surgically.

A 70-year old female initially developed left-dominant chorea-ballism with typical diabetic striatopathic alterations on imaging during a nonketotic hyperglycemic state resulting from unknown type 2 diabetes. Three years later, after spontaneous improvement but not complete resolution of her left-sided chorea-ballism, she developed severe right-sided chorea-ballism following a new hyperglycemic state with improper antidiabetic medication compliance. She underwent a left-sided pallidotomy with immediate benefit and a favorable outcome on both the chorea-ballism and the quality of life after 8 months.

A 64-year old female who developed right-sided chorea-ballism after a hyperglycemic episode due to unknown type 2 diabetes was treated three years later with left pallidal deep brain stimulation (DBS), with good short-term response.



Although pallidotomies and pallidal DBS have been used to treat hyperkinetic movement disorders for 70 and >20 years respectively, these patients are amongst the first to undergo pallidal RF ablation and DBS in post-hyperglycemic chorea-ballism. Contrary to previous descriptions in post-hyperglycemic and post-stroke chorea-ballism, pallidal neuronal firing rates were similar to those recorded in patients with Parkinson’s disease.



We present patients undergoing pallidal RF ablation and DBS for post-hyperglycemic chorea-ballism, with excellent short-term results, suggesting a possible role in medication-refractory cases.

Philippe DE VLOO (Leuven, BELGIUM), Robert DALLAPIAZZA, Darrin LEE, Luka MILOSEVIC, William HUTCHINSON, Alfonso FASANO, Renato MUNHOZ, Anthony LANG, Suneil KALIA, Andres LOZANO
16:05 - 16:10 #16230 - O119 Directional deep brain stimulation of the subthalamic nucleus: towards defining the best directionality and anatomical stimulation site.
O119 Directional deep brain stimulation of the subthalamic nucleus: towards defining the best directionality and anatomical stimulation site.

Background: Directional deep brain stimulation of the subthalamic nucleus for treatment of Parkinson´s disease offers new possibilities of current steering towards clinically effective locations and thereby increasing the therapeutic window. The objective of the current work is to understand the relation between the anatomical stimulation site of directional stimulation and clinical effects.

Methods: Postoperative clinical mapping after 6 months was performed in a prospective, consecutive series of 28 patients. Stimulation effect and side effect thresholds of DBS lead contacts were systematically assessed using the UPDRS motor subscores and compared to baseline in the medication-off state. DBS lead positions and their corresponding volume of tissue activation (VTA) were normalized into the MNI space and integrated into a three-dimensional human brain atlas to construct a stimulation map. Non-parametric statistical tests were applied to test for differences between subgroups.

Results: In 24 of 56 (43%) analysed hemispheres directional stimulation provided a larger therapeutic window compared to omnidirectional stimulation. The mean therapeutic window of the best directional contact was increased by 6.8 ± 60% compared to omnidirectional stimulation whereas it was significantly decreased by 33.2 ± 84% for the worst directional contact (p< 0.001). Clinically effective contacts and their associated VTA projected onto the dorsolateral aspect of the normalized STN. Electrode contacts with a low side effect threshold (< 3.0 mA) projected onto the dorsolateral STN and pointed posterolateral towards the internal capsule.

Conclusion: According to our postoperative mapping results, directional stimulation provides a slightly increased therapeutic window compared to omnidirectional stimulation. The dorsolateral STN seems to be the optimal stimulation site, although stimulation of this subregion is also associated with a low side effect threshold. Follow-up data of chronic stimulation is needed to confirm these preliminary results.

16:10 - 16:15 #16275 - O120 Perceived intraoperative stress of movement disorder patients undergoing awake deep brain stimulation surgery.
O120 Perceived intraoperative stress of movement disorder patients undergoing awake deep brain stimulation surgery.

Background: Patients with movement disorders undergoing implantation of deep brain stimulation (DBS) electrodes are typically awake during surgery. Albeit commonly referred to as a highly stressful event, little is hitherto known about how the conscious patient perceives the different stages of awake DBS surgery and which preoperative and perioperative factors shape this intraoperative experience.

Objectives: Here, we set out to identify key factors determining the subjectively perceived intraoperative level of psychological distress in patients with different movement disorders undergoing awake bilateral microelectrode-guided implantation of DBS electrodes at our institution (Department of Neurosurgery, University Medical Center Hamburg-Eppendorf).

Methods: Based on a pre-defined set of constituent surgical phases, patients were asked to report their current level of distress on a verbally administered numerical rating scale from 0 (equal to no stress) to 10 (equal to worst possible stress) throughout the whole surgical procedure. In addition, heart rate and arterial blood pressure were monitored as indices of autonomic stress. A reference group of 67 patients with Parkinson´s disease (PD; mean age, 62 years; mean duration of disease, 12.8 years; average Hoehn & Yahr stage, 2.8) undergoing bilateral DBS of the subthalamic nucleus in the OFF medicated state was compared with 12 essential tremor patients (mean age, 64 years; mean disease duration, 18 years) and 11 dystonia patients (mean age, 56 years; mean disease duration, 15.5 years) undergoing DBS surgery in the ventrointermediate thalamic nucleus and internal globus pallidus, respectively. We employed linear mixed modelling to test for the effects of several preoperative demographic (e.g., gender and age at surgery) and clinical variables (e.g., disease duration and -severity) and perioperative factors (such as surgical phase, duration of surgery, analgosedation) on the dependent variable ‘level of distress’.

Results: Across all patient groups, the average distress score during awake bilateral DBS surgery generally ranged between 4 and 5 on the numerical rating scale from 0 to 10. PD patients reported significantly higher stress levels compared to patients with essential tremor and dystonia (p = 0.012). Subjectively perceived distress levels were critically dependent on surgical phase (p < 0.001), clearly peaking during periods of test stimulation with high frequency (130Hz; pulse width, 60µs). Furthermore, titration of analgosedation with low-dose remifentanil had a significant, time-delayed, effect (p = 0.038). The subjectively perceived distress level did not depend on any other factors including sex, age at surgery, disease duration and duration of surgery (all p > 0.05). The factor ‘disease duration’ was still non-significant (p = 0.081) following separate mixed-effects modelling for the group of parkinsonian patients. However, a median split of the PD sample revealed that patients with longer disease history (>11 years) reported higher distress levels than those with shorter disease duration, presumably paralleling differential motor deterioration following presurgical levodopa withdrawal. Objective stress measures (heart rate, blood pressure) were highly and positively correlated with the reported distress scores.

Conclusions: The verbally administered numerical distress rating scale is a simple and practical tool to monitor the patient´s distress level throughout the course of awake DBS interventions. Our results suggest that distress level is maximal during surgical steps that require the active cooperation of the patient, such as intraoperative test stimulation for unwanted side effects. Furthermore, our results also show that peak stress periods during awake DBS surgery can effectively be countered by careful titration of analgosedation with remifentanil.

Johanna SIEGER, Alessandro GULBERTI, Johannes KOEPPEN, Hans O PINNSCHMIDT, Rainer NITZSCHKE, Miriam SCHAPER, Carsten BUHMANN, Andreas K ENGEL, Anja MEHNERT, Christian GERLOFF, Manfred WESTPHAL, Monika POETTER-NERGER, Wolfgang HAMEL, Christian Ke MOLL, Christian Ke MOLL (Hamburg, GERMANY)
16:15 - 16:20 #16276 - O121 Assessment of directional deep brain stimulation in essential tremor.
O121 Assessment of directional deep brain stimulation in essential tremor.

Background and Objective     

Deep brain stimulation (DBS) applied with directional leads within the ventrolateral thalamus and subthalamic region may lead to differential effects on tremor suppression, and it may alter the threshold for eliciting paraesthesias.



Nine patients suffering from essential tremor (mean age 68,5 yrs, average disease duration 23 yrs) were implanted uni- (1) or bilaterally (8) with directional leads for deep brain stimulation of the ventrolateral thalamus and subthalamic region. Acute effects elicited by stimulation in different directions via bipolar activation of corresponding segments pointing into the same direction (lower segment = cathode) were assessed intraoperatively. Monopolar review (stimulation with 60 usec, 130 Hz) was performed after postoperative microlesioning effects (mean 240 days) had vanished. This involved clinical ratings of tremor suppression, including quantitative accelorometer measurements, and the assessment of the threshold and intensitiy of paraesthesias. The patients and raters were blinded with regard to the direction of stimulation that was chosen in random order and stimulation amplitude.



None of the patients experienced adverse events related to the surgical procedure or the implanted hardware. All patients exhibited a microlesioning effect and sustained tremor suppression with chronic stimulation. Intraoperative assessments revealed a lower threshold for paraesthesia when stimulation was performed into the posterior (medial and lateral) direction compared to stimulation into the anterior direction. Monopolar review revealed that thresholds for tremor suppression and paraesthesias (all transient) were very variable among patients, and for three electrodes paraesthesia thresholds did not differ between directions. For 73% of the contacts, directional stimulation via segments increased the therapeutic window compared to ring mode stimulation, in particular, if stimulation was applied into the anterior and medial direction. The maximum difference between stimulation amplitudes required for tremor suppression and those eliciting paraesthesias ('therapeutic window') was 2 mA. Based on these assessments, in four patients directional stimulation was commenced and rated superior by the patients at follow-up visits.


The influence of directional DBS on tremor suppression and paraesthesia thresholds is highly variable. Although stimulation in the anterior and medial direction appeared superior in several cases, a uniform pattern among patients could not be observed. Clinical responses may depend on the location of contacts, individual anatomy and biophysical characteristics of directional DBS that required further investigation. 

Miriam SCHAPER (Hamburg, GERMANY), Christian Ke MOLL, Alessandro GULBERTI, Ute HIDDING, Carsten BUHMANN, Andreas Ak ENGEL, Christian GERLOFF, Manfred WESTPHAL, Johannes A KOEPPEN, Monika POETTER-NERGER, Wolfgang HAMEL
16:20 - 16:25 #16287 - O122 Phase resetting mechanism of thalamic oscillatory activity contributes to the generation of the somatosensory evoked potentials in Vim thalamus in patients with essential tremor and Parkinson’s disease.
O122 Phase resetting mechanism of thalamic oscillatory activity contributes to the generation of the somatosensory evoked potentials in Vim thalamus in patients with essential tremor and Parkinson’s disease.


    In the past few years it has become increasingly acknowledged that large scale oscillatory activity in sensorimotor system plays an important role in physiological brain function as well as pathophysiology of neurological disorders. Thalamic oscillations were commonly reported in Parkinson’s disease (PD). On the other hand, previous studies demonstrated median nerve stimulation evoked potentials in the nucleus ventralis intermedius (Vim) of the thalamusin PD patients (Hanajima R et al, 2004). In fact, the Vim is believed to receive kinesthetic projections including muscle afferent projections (Ohye et al., 1989).

    In the first conventional view, the stimulus presentation evokes new event-related neural activity strictly time-locked to the stimulation which is superimposed on the background activity. And response averaging removes background activity (considered to be noise), whose time course is presumed to be independent of median nerve stimulation. Recently, by contrast, SEP features arise from alterations in the dynamics of ongoing neural synchrony generating thalamic LFP. By these accounts, SEP features are produced through stimulus-induced phase resetting of ongoing field potential oscillations, a phenomenon observed in vitro.

    To address this question, we investigated whether reorganization of background thalamic oscillatory activity contributes to the generation of the evoked potentials in Vim thalamus triggered by median nerve stimulationin patients with essential tremor and Parkinson’s disease.


    We recorded mediannerve stimulation-elicited somatosensory evoked potentials (SEPs) from the ventralis intermedius (Vim) thalamus with semi-microelectrodes during stereotactic surgery in 10 patients with Parkinson’s disease (n=3) and essential tremor (n=7). Hilbert transform was used to measure the phase of the thalamic oscillatory SEP responses for each frequency band. Then we here calculated the phase-locking factor (PLF) in order to measure the phase correlation of ongoing local field potentials (LFPs) oscillation across trials for given frequency bands in the generation of the SEPs in Vim thalamus. This measure corresponds to the inter-trial coherence, which indexes the degree of phase synchronization of trials relative to stimulus presentation. The PLF measure takes values between 0 (absence of synchronization) and 1 (perfect synchronization) (Strogatz, 2000). 


    In about two thirds of thalamic SEPs examined, we found phase-locking of γ frequency band comes first in temporal order, and subsequently that of β band, and followed by phase-locking of slower frequency band (θ and α) of ongoing thalamic LFP oscillation across trials.Transition across trials from uniform to packed phase distribution revealed temporal phase reorganization of given rhythms of thalamic LFP oscillation in relation to stimulation. Otherwise, exclusive slower frequency oscillations such as θ and α band were observed at late period in some SEPs, while no obvious higher frequency band appeared at early period. In some of the surrounding LFPs 3mm apart from those of central electrode, exclusive slower frequency oscillations such as θ and α band were observed at late period, while no obvious higher frequency band was present at early period.


    The identification of a phase-locking in each rhythm of thalamic SEP trials reinforces the contribution of phase resetting model for somatosensory information processing. This may imply that phase resetting of each ongoing LFP rhythm in an appropriate temporal order at least partly participates in thalamic SEP formation. Understanding characteristic transient responses of neuronal populations to external stimulations may provide us with invaluable clues for investigating how the central nervous system such as thalamic nucleus reacts and adapts to the stimulus perturbations. For the therapeutic applications, however, it is a great challenge to design deep brain stimulation (DBS) techniquesas effective as possible.

Katsushige WATANABE (Tokyo, JAPAN), Sumito SATO, Yasushi OKAMURA, Hiromi KAMO, Makoto TANIGUCHI
16:25 - 16:30 #16295 - O123 Investigating the effect of STN-DBS stimulation and different frequencies settings on the acoustic-articulatory features of vowels.
O123 Investigating the effect of STN-DBS stimulation and different frequencies settings on the acoustic-articulatory features of vowels.


In Parkinson’s disease (PD) in addition to motor symptoms, nonmotor symptoms and voice and speech disorders can also develop in 90% of PD patients. The aim of our study was to investigate the effects of DBS and different DBS frequencies on speech acoustics of vowels in PD patients.


The study included 16 patients who underwent STN-DBS surgery due to PD. The voice recordings for the vowels including [a], [e], [i], and [o] were recorded at frequencies including 230 Hz, 130 Hz, 90 Hz, 60 Hz and off-stimulation. The voice recordings were evaluated by the Praat software and the effects on the first (F1), second (F2), and third formant (F3) frequencies were analyzed.


A significant difference was found for the F1 value of the vowel [a] at 130 Hz compared to off-stimulation. However, no significant difference was found between the three formant frequencies with regards to the stimulation frequencies and off-stimulation. In addition, though not statistically significant, stimulation at 60 Hz and 230 Hz led to several differences in the formant frequencies of other three vowels (Table-1).


Our results indicated that STN-DBS stimulation at 130 Hz had a significant positive effect on articulation of [a] compared to off-stimulation. Although there isn’t any statistical significant stimulation at 60 Hz and 230 Hz may also have an effect on the articulation of [e], [i], and [o] but this effect needs to be investigated in future studies with higher numbers of participants.

Atilla YILMAZ (Hatay, TURKEY), Elif Tugba SARAC, Fatma Esen AYDINLI, Mustafa Turgut YILDIZGOREN, Esra OKUYUCU, Mustafa ARAS, Şükrü ORAL, Yurdal SERARSLAN
16:30 - 16:35 #16320 - O124 Outcome of Subthalamic Nucleus deep brain stimulation in Parkinson’s Disease after 15 years.
O124 Outcome of Subthalamic Nucleus deep brain stimulation in Parkinson’s Disease after 15 years.

Background and objectives:

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) is a proven effective and safe intervention to treat motor symptoms in Parkinson’s disease (PD). Little is known about long-term follow-up (>10 years) after implantation. Our goal is to provide an analysis of motor and cognitive outcome of  PD patients 15 years after surgery for STN-DBS. Additionally, we analysed the burden of the caregiver. To limit the loss to follow-up, patients unable to visit the hospital have been assessed at their homes.


In this observational study, we report on the motor/cognitive outcome and caregiver burden in a cohort of 15 PD patients who underwent STN-DBS in the Maastricht University Medical Centre (MUMC+) from January 1999 to December 2003. Motor outcome has been assessed by the Unified Parkinson's Disease Rating Scale (UPDRS) part III and cognition by the Mini-Mental State Examination (MMSE). The caregiver burden was assessed by the Zarit, without assessment before surgery. The UPDRS I and II were used to assess cognition and activities of daily life and the VAS EuroQol (EQ)5D for perception of health.


Between 1999 and 2003 43 patients were operated of which 15 patients survived follow-up. Motor function (UPDRS III) significantly worsened compared to baseline with antiparkinsonian medication (47.7 vs. 18.3, p=0.005) and was comparable to baseline without antiparkinsonian medication UPDRS III (44.7). Axial symptoms and bradykinesia are most severely affected of all motor symptoms at 15 years after surgery.

At 15 years after surgery, 40% of patients (n=6) met the diagnostic criteria of dementia and PD Mini-mental state examination (MMSE) scores non-significantly declined (baseline 28.0, 15 year FU 22.7). UPDRS II significantly worsened (p<0.01; BWM 10.5, 15 year FU 30.5). UPDRS I significantly declined (p<0.005; BWM 1.46, 15 year FU 6.7)

VAS EQ5D scores were 0.55, EQ-5D scores were 0.43 and 62% (n=8) had signs of mild mood disturbances to severe depression at FU.

A Zarit score above 16, suggestive of clinically significant caregiver burden, was reached by one third (n=5) of caregivers at 15 years after surgery. Mean Zarit scores were 17.9±9.4 and 10±4.9 for caregivers of patients living in their own residence (n=8) and in nursing homes/retirement homes (n=7), respectively, however no significance was found.


Cognitive and axial symptoms are highly dominating after 15 years of follow-up, whereas tremor and rigidity show less deterioration over time. This has a significant impact on health status, quality of life and caregiver burden.

Lucas Guido WESTERINK (Rotterdam, THE NETHERLANDS), Yasin TEMEL, Linda ACKERMANS, Annelien DUITS
16:35 - 16:40 #16331 - O125 The Impact of Microelectrode Recording on Lead Placement Accuracy During Deep Brain Stimulation Surgery.
O125 The Impact of Microelectrode Recording on Lead Placement Accuracy During Deep Brain Stimulation Surgery.


The effect of microelectrode recording (MER) on DBS placement is not well understood.  In order to assess the impact of MER on final DBS lead location in relation to the original anatomical target selected, we compared radial error of DBS lead coordinates in relation to initial target in two groups of hemispheres –Hemispheres whereby the MER resulted in an intentional deviation of DBS placement; versus hemispheres whereby MER did not result in any change in DBS placement location.  


Hemispheres of patients who underwent DBS surgery with MER at the authors’ institution from 2014-2018 were retrospectively analyzed. Targets consisted of the subthalamic nucleus (STN), ventral intermediate nucleus of the thalamus (VIM), internal globus pallidus (GPi) and zona incerta (ZI). The hemispheres were subdivided into two groups: those in which the initial MER track was chosen for final lead implantation and those where a different track from the initial was chosen. Coordinates of the intended MER-based target and final DBS lead were calculated within the same z-axis plane using intraoperative computed tomography (CT) merged to preoperative magnetic resonance imaging (MRI) studies.  Radial error between MER-based target and final DBS lead was then calculated in both groups and compared using a two-tailed T-test.  


Two hundred forty-seven hemispheres were reviewed (157 STN, 57 VIM, 31 GPi, 2 ZI).  There were 96 hemispheres where the initial MER track was chosen for lead implantation and 151 where a different track was chosen.  Radial error of the DBS lead in hemispheres with lead implantation into the initial MER track was significantly (p<0.05) less than hemispheres where a different track was chosen for lead implantation (0.84±0.07 vs 1.1±0.06 mm).  Radial error broken down by target is shown in Table 1. 


There was a significantly higher radial error of the final DBS lead in hemispheres where a different track from the initial MER track was chosen.  While the difference reached statistical significance, the numerical value of the difference was small.

16:40 - 16:45 #16342 - O126 Corticospinal Tract Activation Threshold Varies in Ring versus Segmented Mode Stimulation in DBS Surgery Using Motor Evoked Potential Measurements.
O126 Corticospinal Tract Activation Threshold Varies in Ring versus Segmented Mode Stimulation in DBS Surgery Using Motor Evoked Potential Measurements.

Corticospinal Tract Activation Threshold Varies in Ring versus Segmented Mode Stimulation in DBS Surgery Using Motor Evoked Potential Measurements

Ryan B. Kochanski MD, Jay Shills, PhD, Gian D. Pal MD MS, Leo Verhagen MD PhD, Sepehr Sani MD


The introduction of segmented leads in DBS surgery promises to improve current delivery by directing current in an intended direction and perhaps more importantly, by avoiding current activation of undesired local anatomy.  The extent to which segmented stimulation away from the corticospinal tract (CST) allows for increased side effect threshold as compared to stimulation in ring mode is not well studied. 


CST activation thresholds were evaluated using motor evoked potentials (MEP).  Electrodes were placed in the contralateral muscles of the face and upper extremity for MEP measurements.  Stimulation (unipolar) was delivered by delivering current to inserted DBS leads (20 hemispheres total; 10 STN, 10 VIM) intra-operatively.  Stimulation parameters included activation of each segment of the middle two contacts followed by ring mode stimulation of all four contact while records MEP thresholds and muscle activated. 


MEP thresholds were obtained from all stimulated segments and rings.  In VIM, the segment with highest threshold was noted at current that was on average 1.1mA above that in ring mode.  The abductor pollicis brevis muscle was the first MEP observed.  In STN stimulation, the results were mixed.  Variable muscle MEPs were seen at threshold stimulation.  Segmented stimulation thresholds were variable when compared to ring mode with the difference in current threshold varying between 0.6 to 1.2mA.  A likely explanation is lack of segment alignment directly opposite of CST along with the curved trajectory of CST superior and lateral to STN.


Segmental DBS stimulation leads to mild increase in CST activation threshold in VIM and STN.  Increased thresholds can be variable in STN likely due to imperfect segmental alignment as well as local anatomy of CST.

Ryan KOCHANSKI, Jay SHILLS, Leonard Verhagen METMAN, Sepehr SANI (Chicago, USA)
16:45 - 16:50 #16345 - O127 Activation of pallidofugal and nigrofugal fibers with effective subthalamic nucleus deep brain stimulation for Parkinson’s disease.
O127 Activation of pallidofugal and nigrofugal fibers with effective subthalamic nucleus deep brain stimulation for Parkinson’s disease.


Deep brain stimulation of the subthalamic nucleus (STN) is an effective therapy for patients with Parkinson’s disease (PD). Different regions within the subthalamic area have been correlated with optimal clinical outcomes. In this work, we correlate the motor outcome in PD patients with the involvement of nigrofugal and pallidofugal pathways by effective stimulation.


Fourty-three patients with STN-DBS were included and their clinical and stimulation parameters were recorded at one-year follow-up. Nigrofugal and pallidofugal pathways were obtained using constrained spherical deconvolution probabilistic tractography with imaging data from 43 PD matched subjects from the Parkinson's Progression Markers Initiative connectome. The volume of activated tissue (VAT) for all patients was modelled using finite element method within the LeadDBS software. All VATs were summed and analyzed using generalized linear models in FSL software to obtain statistically significant stimulation clusters correlated with: greater reduction of dopaminergic medication, improvement in Unified Parkinson's Disease Rating Scale (UPDRS) III, bradykinesia, rigidity, and tremor scores. Finally, we calculated the overlap coefficient of these clusters with the nigrofugal or pallidofugal fibers.  Then, these coefficients were compared with the overlap of the significant clusters with the STN.


Significant reduction of symptoms was obtained with STN-DBS in our patients (p<.05). Nigrofugal and pallidofugal pathways were traced as described in previous anatomical descriptions. Significant clusters associated with the greatest dopaminergic medication reduction, greatest UPDRS improvement overlapped with the nigrofugal fibers (coeff0.50) but not with the STN. Significant clusters associated with greater bradykinesia improvement overlapped with the nigrofugal fibers (coeff 0.55) and with the STN (coeff 0.19). Finally, significant clusters associated with greater rigidity and tremor improvement overlapped with pallidofugal fibers (coeff 0.60) and not with the STN.


Electrical field involvement of the nigrofugal fibers, more than the STN, appears to produce the most improvement in UPDRS III, bradykinesia, and the most reduction of dopaminergic medications. Similarly, involvement of the pallidofugal fibers and not the STN was associated with greater tremor and rigidity improvement.

Josue AVECILLAS-CHASIN (Tarragona, SPAIN), Christopher HONEY
16:50 - 16:55 #16355 - O128 Increasing the proportion of rechargeable neurostimulators: Estimated impact on the financial sustainability of a deep brain stimulation program.
O128 Increasing the proportion of rechargeable neurostimulators: Estimated impact on the financial sustainability of a deep brain stimulation program.

Background: Deep brain stimulation (DBS) has been shown to be cost-effective in the treatment of Parkinson's disease. However, it represents a costly burden for national health systems. This situation has been aggravated by the the reduced longevity of newer non-rechargeable batteries, which has led to a rise of battery replacements costs and complications (i.e., device infections), somewhat compromising the sustainability of DBS programs. We hypothesized that investing on rechargeable devices in the short term would result in a reduction of replacements and, hence, of total DBS costs in the middle-to-long term.

Objective: 1) To estimate deep brain stimulation (DBS) material costs for a 6-year-period at our institution considering two scenarios: a conservative one focused on gradual increase of new rechargeable devices; another one with a greater initial investment in rechargeable devices (both new cases and replacements); 2) to compare these estimates to costs of the preceding three years (2014-2016), when devices had higher prices and were mostly non-rechargeable.

Methods: Descriptive analysis of the activity and material costs of DBS at our institution in 2014-2016; estimation of DBS material costs in 2017-2022 based on the aforementioned scenarios; calculation of the shortfall vs. saving resulting from comparison of the different DBS cost estimates to the costs from the 2014-2016 period.

Results: During the 2014-2016 period, the total DBS cost rose from 654.958,14€ in 2014 to 1.023.689,64€ in 2016 (56,30% increase). Regarding the estimates for the 2017-2022 period, the first scenario would benefit initially from devices price-cuts dropping to 907.665,00€ in 2017 (-11,30% compared to 2016, yet +38.58% relative to 2014), but would eventually resume the upwards trend, ending up in 2022 with a total DBS cost of 1.058.541,00€ (+3,40% vs. 2016; +61,62% vs. 2014). Conversely, the second scenario, in spite of greater cost-increase the first 3 years, would eventually lead to a reduction of replacements and subsequently of total DBS costs down to 775.676,00€ in 2022 (-24,03% relative to 2016; only +18,43% compared to 2014).

Conclusion: The cost and saving estimates of increasing the proportion of implanted rechargeable DBS neurostimulators versus primary cell devices (both new cases and replacements) support the notion that this strategy, while costly in the short term, could aid to ensure the sustainability of DBS programs in the long run.



[1]         Eggington S, Valldeoriola F, Chaudhuri KR, et al. The cost-effectiveness of deep brain stimulation in combination with best medical therapy, versus best medical therapy alone, in advanced Parkinson's disease. J Neurol 2014; 261: 106-16.

[2]         Pepper J, Zrinzo L, Mirza B, et al. The risk of hardware infection in deep brain stimulation surgery is greater at impulse generator replacement than at the primary procedure. Stereotact Funct Neurosurg 2013; 91: 56-65.

16:55 - 17:00 #16359 - O129 Effects of spinal cord stimulation on postural control in Parkinson's disease patients with freezing of gait.
O129 Effects of spinal cord stimulation on postural control in Parkinson's disease patients with freezing of gait.

Background:Freezing of gait (FoG) in Parkinson’s disease (PD) is an incapacitating transient phenomenon, followed by continuous postural disorders. Spinal cord stimulation (SCS) is a promising intervention for FoG in patients with PD, however its effects on distinct domains of postural control is not well known.Aim:The aim of this study is to assess the effects of SCS on FoG and distinct domains of postural control. Methods:Four patients with FoG were implanted with SCS systems in the upper thoracic spine. Anticipatory postural adjustment (APA), reactive postural responses, gait and FoG were biomechanically assessed. Results:In general, the results showed that SCS improved FoG and APA. However, SCS failed to improve reactive postural responses. SCS seems to influence cortical motor circuits, involving the supplementary motor area. On the other hand, reactive posture control to external perturbation that mainly relies on neuronal circuitries involving the brainstem and spinal cord, is less influenced by SCS. In this short report we provide data suggesting that SCS may affect important measures in anticipation of gait that decreases the chance of FoG in gait initiation. On the other hand, our data provides evidence that SCS failed to improve reactive balance control. We suggest that SCS has distinct effects on specific postural control domains, which is important to develop and improve neuromodulation-based interventions to treat postural and gait disabilities in PD. 


Figure 1 – (A) Characterization of FoG based on spectral analysis. Schematic representation of a man representing the lumbar accelerometer as a small black box. The curve represents the vertical acceleration acquired during step initiation. The dotted square shows a 7.5s window for the frequency analysis domain (normal gait in blue and freezing in red). (B) Spectral analysis of acceleration showing one band representing the locomotor period (0 - 3Hz) and the FoG band (3 - 8Hz). (C) FoG index showing the clinical threshold (>2=FoG); FoG index is calculated by dividing the FoG band by the locomotor band (blue circle – power peak of normal gait; red circle – power peak of a period with FoG). (D) Representation of the step initiation task, showing the marker on the malleolus to detect the moment that the foot clears the floor. The sequence showed in D and E (1-4) shows: (1) beginning of the task, when the participant is in quiet standing (body weight balanced under the feet), preceding the step. (2) The red arrow shows body weight shifting toward the supporting leg (APA). (3) Once the body weight is shifted contralaterally, the moving leg can be released to take a step and finish the movement (4). The graphs showed in D and E represent the moment that each phase of the sequence (1-4) occurs: 1-quiet standing; 2- peak of APA; 3- step; 4-end of the movement. The red line represents the mediolateral force under the supporting leg, the dotted line is the displacement of the moving foot. The red line in D represents a defective mediolateral displacement of the body weight (APA), showing a relative smaller amplitude and longer APA compared to a normal APA displayed in E.

Figure 2 – Individual and mean values for % change of the outcome variables: time of gait, time of FoG, time of APA, amplitude of APA and the variables that describes reactive postural control - CoPap amplitude and CoMap amplitude. (A) Individual mean curves of the body mediolateral displacement (APA) - thin lines; and the standard deviation (thick and transparent bands) for each stimulation parameter (OFF – gray, 60 Hz – red, 300 Hz – blue). (B) Graphs showing the individual mean values of % of change for each outcome variable in 60 and 300 Hz relative to the off condition. (C) Mean values and standard deviations of % change for each outcome variable (time of gait, time of FoG, time and amplitude of APA, CoP and CoM amplitude. Asterisks highlight significant effects for 60 and 300 Hz as compared to the off condition.

Daniel Boari COELHO, Carolina DE OLIVEIRA SOUZA (São Paulo, BRAZIL), Erich FONOFF, Clement HAMANI, Dos Santos Ghilardi MARIA GABRIELA, Luis Augusto TEIXEIRA

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
15:00 - 17:00

Parallel Session 13
Animal Models & Basic Science

Moderators: Hagai BERGMAN (Prof) (Jerusalem, ISRAEL), A Holly ROY (UK)
15:00 - 15:15 PPN in monkey. Carine KARACHI (MEDECIN) (PARIS, FRANCE)
15:15 - 15:30 Animal model for Schizophrenia - first steps towards closed-loop DBS treatment. Hagai BERGMAN (Prof) (Jerusalem, ISRAEL)
15:30 - 15:40 #16356 - O130 Higher endogenous alpha-synuclein confers greater disease susceptibility in parkinson’s; clinical-grade alpha-synuclein null human embryonic stem cells may be used to create disease-resistant dopaminergic grafts.
O130 Higher endogenous alpha-synuclein confers greater disease susceptibility in parkinson’s; clinical-grade alpha-synuclein null human embryonic stem cells may be used to create disease-resistant dopaminergic grafts.


Open-label trials in the 1990s showed that some Parkinson’s disease (PD) patients had marked improvement in motor function following ventral mesencephalic fetal graft transplantation into striatum. However, decades later on autopsy the grafts acquired Lewy pathology, a hallmark of PD, with concomitant loss of graft function. Focus is now moving towards human embryonic or induced pluripotent stem cell-derived grafts to overcome limitations of using fetal tissue, but host-to-graft spread of disease remains a major issue for these future treatments. We developed an in vitro model of disease spread using transgenic human embryonic stem cell (hESC) lines over-expressing alpha-Synuclein, the protein responsible for Lewy pathology. The aim of this work is to investigate the importance of endogenous alpha-Synuclein expression on disease susceptibility and generate disease-resistant dopaminergic grafts to test host-to-graft spread in vivo.


The Shef4 hESC line was transfected with an expression plasmid containing the human SNCA gene (and CAG promoter) to produce several clonal transgenic lines over-expressing alpha-Synuclein. Of these, the selected S37 and S8 clonal lines expressed eight-fold greater levels of α-synuclein than the parental line. These three transgenic hESC lines and an alpha-Synuclein knockout hESC (RC17) line were differentiated into cortical neurones (also affected by Lewy pathology in PD). These neurones were then seeded with recombinant monomers and pre-formed fibrils (PFFs) of alpha-Synuclein for a week. After a further 3 weeks, the neurones were fixed and stained to detect the expression of phosphorylated alpha-Synuclein at Serine-129 (pSer-129), an early marker of Lewy pathology.


Monomers of alpha-Synuclein did not induce the expression of pSer-129 alpha-Synuclein in any hESC line. PFFs led to the formation of pSer-129 alpha-Synuclein structures in S37, S8 and the control Shef4 line but not in the alpha-Synuclein knockout line. The pSer-129 alpha-Synuclein immunostaining was significantly higher in S37 and S8 relative to Shef4 (p<0.001). Immunostaining was observed as short discrete axonal as well as perinuclear structures resembling Lewy body and neurite formation.


We show that monomeric alpha-Synuclein cannot seed Lewy body-like pathology, whereas PFFs can initiate such pathology in human cortical neurones. This is in agreement with published data on rodent models. We show that higher endogenous expression of alpha-Synuclein confers greater disease susceptibility and conversely the alpha-Synuclein knockout line does not seed any pathology. This work establishes an important human model to study PD and alpha-Synuclein knockout hESC line-derived dopaminergic grafts have been generated to test in vivo. Future experiments will aim to show such grafts can rescue the phenotype in classic lesion models of PD and that they may be resistant to PFF-induced Synucleinopathy.


Ammar Natalwala is funded by the Wellcome Trust Research Training Fellowship.

Ammar NATALWALA (Edinburgh, UK), Yixi CHEN, Karamjit DOLT, Ratsuda YAPOM, Marco KRIEK, Terry BAKER, Patrick DOWNEY, Tilo KUNATH
15:40 - 15:45 #16140 - O131 Dorsolateral striatal neuronal activity in a rat model of Parkinson`s disease with levodopa-induced dyskinesias.
O131 Dorsolateral striatal neuronal activity in a rat model of Parkinson`s disease with levodopa-induced dyskinesias.


Striatal inhibitory interneurons are important modulatory components for the direct and indirect pathways of basal ganglia (BG). However, their role in Parkinson's disease (PD) and PD with levodopa induced dyskinesias (LID) remains unclear.


Single neuron activity and local field potentials (LFPs) were recorded from the dorsolateral striatum (DLStr) together with electrocorticogram (ECoG) from the motor cortex (MCtx) area in 6-hydroxydopamine lesioned hemiparkinsonian (HP) and levodopa-primed dyskinetic (HP-LID) rats as compared to controls. Putative GABAergic interneurons (fast-spiking interneurons, FSIs) and medium spiny neurons (MSNs) were identified based on peak to trough. For the calculation of beta (13-30Hz) and gamma (30-100Hz) bursts the signals were filtered using Blackman bandpass filter and further down sampled to 100 Hz. Hilbert transformation was used to get the envelope and threshold value for burst was set at the 75th percentile of signal amplitude.


The MSNs firing rate was enhanced in HP and HP-LID rats, while firing rate in FSIs were only enhanced in HP-LID rats together with reduced burst activity. The coefficient of variation (measure of firing irregularity) was only higher in HP rats in both MSNs and FSIs, with no difference in HP-LID rats. Only in HP-LID MSNs spike-triggered averaging of LFP were phase locked to beta (12-30 Hz) and gamma (30-100 Hz), while the effect in HP was less pronounced. Further, ECoG oscillatory activity in the beta and gamma burst count was enhanced in both HP and HP-LID rats, however, theta burst was higher only in HP-LID rats.


Our study suggests that there may be a distinct form of MSNs and FSNs firing properties and MCtx burst activity in a rat model of PD with dyskinesias, which may depend on the functional state after dopamine depletion and treatment indicating maladaptive neuroplasticity.

Xingxing JIN, Denny MILAKARA, Kerstin SCHWABE, Joachim K KRAUSS, Mesbah ALAM (Hannover, GERMANY)
15:45 - 15:50 #16149 - O132 Regulation of viral vector gene expression in the hippocampus by medial septal nucleus deep brain stimulation.
O132 Regulation of viral vector gene expression in the hippocampus by medial septal nucleus deep brain stimulation.


Deep brain stimulation (DBS) is a well-established treatment for several neurological diseases; however, the mechanisms by which DBS exerts its effects remain controversial. Electrical stimulation has clear effects on nearby cells but can also affect distant neuronal populations within the same circuitry. This proof-of-principle study aims to examine whether DBS can regulate gene expression from a virally delivered vector by first transducing neurons with an adeno-associated virus containing the gene for a fluorescent reporter protein downstream of a DBS-responsive promoter and then stimulating with DBS from a remote target.


Adult Sprague-Dawley rats underwent a unilateral right hippocampal injection of adeno-associated virus serotype 5 containing a DNA construct for green fluorescent protein (GFP) under control of the chicken beta actin promoter and TdTomato (TdT) under control of the c-fos promoter  (AAV5-c-fos). Two weeks later, a stimulating depth electrode was implanted into the medial septal nucleus (MSN). In the first experiment, 1 week after implantation, rodents received no DBS, 7.7 Hz (theta frequency) DBS, or 130 Hz (gamma frequency) DBS for 1 hour, and tissue analysis was performed 1.5 hours after conclusion of stimulation. In the second experiment, rodents received DBS treatment 1 week after implantation followed by another treatment 1 week after the initial treatment with no DBS, 7.7 Hz DBS, or 130 Hz DBS. Fluorescent protein expression levels were analyzed with confocal microscopy to evaluate for spatio-temporal control of the reporter’s expression in the hippocampus.


In the first experiment, we found that, two weeks after AAV5-c-fos injection, either 1 hour of continuous 7.7 Hz, or 130 Hz MSN stimulation resulted in an increase in TdT reporter expression levels relative to no stimulation. Results from the second experiment demonstrated that TdT levels were no longer increased 1 week later with no additional stimulation, but TdT expression could be induced again with repeat 7.7 Hz or 130 Hz MSN stimulation.


Here, we demonstrate that viral vector-mediated gene expression can be induced using both low- and high-frequency DBS to a distal target along a circuit. Taken together these data suggest that DBS can regulate gene expression both spatially and temporally. Successful control of gene expression by DBS will warrant further investigation into stimulation responsive promoters for use in clinical applications.

Darrin LEE (Toronto, CANADA), Chris MCKINNON, Mitch DE SNOO, Anton FOMENKO, Eun Jung LEE, Victoria AGAPOVA, Sophie NGANA, Clement HAMANI, Andres LOZANO, Lorraine KALIA, Suneil KALIA
15:50 - 15:55 #16386 - O133 The combination of experimental STN DBS and CDNF in rat model of advanced Parkinson's disease.
O133 The combination of experimental STN DBS and CDNF in rat model of advanced Parkinson's disease.

Although, several therapies are effective against the cardinal motor symptoms of Parkinson’s disease (PD), there are no disease course altering therapies currently available [1]. Several neurotrophic growth factors (NTFs) have been effective in animal models of early PD reducing or partially reversing dopaminergic cell death [2]. However, their effect has been limited in clinical trials [3]. However, most clinical NTF trials have been done in patients who have advanced PD where near total dopaminergic cell death has already happened. STN DBS is accepted therapy in advanced PD and it has shown some neuroprotective potential of its own. 

Aim: To study if CDNF can improve the effect of STN DBS in a model of advanced PD.

Methods:  Unilateral 6-OHDA injections in medial forebrain bundle were used to produce a near total dopaminergic depletion, which was verified with amphetamine induced rotations. 

Short-term STN DBS was modeled with unilateral implantation of STN electrodes combined with either CDNF (n = 15) or PBS (n= 11) injected just above substantia nigra pars compacta. The effect of STN DBS was measured by repeated cylinder tests with no stimulation (baseline) and at two stimulation amplitudes. Biochemical measured with HPLC of dopamine (DA) and DA metabolites at 4 weeks or with TH and dopamine transporter (DAT) IHC (at 7 weeks) (Figure 1A).

Long-term STN DBS was modeled with STN lesions (STNL) done with 10 µg ibotenic Acid (IBOT). Group one received PBS + PBS (n=12), group two received CDNF + PBS (n=11), group three received PBS + IBOT (n= 14), and group four received CDNF + IBOT (n=17). The behavioral effect was measured with repeated cylinder tests and apomorphine-induced rotation tests. The biochemical effect was measured by HPLC (4 weeks) or IHC (7 weeks; Figure 2A)

Results: The use of contralateral front limb increased more at 2 and 3 weeks for CDNF-treated animals compared to PBS-treated animals with both low and high stimulation amplitudes when compared to week 1 baselines (Fig 1B and C, no stimulation U = 2.18, p = 0.029, respectively, for low and high stimulation, Fig 1B-C). The optical density of TH-stained striatum was higher in CDNF + STN HFS co-treated rats compared to contralateral side than in PBS and STN-HFS treated animals, (Fig1 E-C). The DA and DA metabolite levels were similar, Fig 1 H and I. 

Only in the combination of CDNF and STNL the use of contralateral front limb use was higher (Fig 2B -C) and the number of apomorphine induced rotations were lower compared to baseline and double-sham PBS + PBS (Fig 2D). There were now differences in biochemical analysis (Fig2E-H).

Conclusions: We found additive but transient synergistic effect between STN DBS and CDNF in behavioral tests in a model of advanced PD where neurorestorative treatments have usually failed. Although there was no biochemical evidence of neuroprotection, CDNF has been previously reported to provide functional benefits without biochemical effects [4].

There are no previous reports on the combination of DBS and neurotrophic factors. This combinatory effect warrants further studies. Additionally, in human brain STN and SNpc are close and could be reached through single or adjacent trajectories. 


Results published in Huotarinen A, Penttinen A-M, Bäck S, et al. Combination of CDNF and Deep Brain Stimulation Decreases Neurological Deficits in Late-stage Model Parkinson's Disease. Neuroscience. 2018;374:250-263. doi:10.1016/j.neuroscience.2018.01.052.





1.        Kalia LV, Kalia SK, Lang AE: Disease-modifying strategies for Parkinson's disease. Mov Disord 2015 Jul 24;30:1442–1450. 

2.        Airavaara M, Voutilainen MH, Wang Y, Hoffer B: Neurorestoration. Parkinsonism Relat Disord 2012 Jan;18:S143–S146. 

3.        Hegarty SV, Lee DJ, O'Keeffe GW, Sullivan AM: Effects of intracerebral neurotrophic factor application on motor symptoms in Parkinson's disease: A systematic review and meta-analysis. Parkinsonism Relat Disord 2017 May;38:19–25. 

4.        Bäck S, Peränen J, Galli E, Pulkkila P, Lonka-Nevalaita L, Tamminen T, et al.: Gene therapy with AAV2-CDNF provides functional benefits in a rat model of Parkinson's disease. Brain Behav 2013 Mar;3:75–88. 

16:00 - 16:05 #16397 - O135 Acute Fornix DBS induces long-term depression of hippocampal synaptophysin levels.
O135 Acute Fornix DBS induces long-term depression of hippocampal synaptophysin levels.

Fornix DBS has the ability to refurbish memory function in animal models with experimental dementia. One of the possible underlying mechanisms is the acute increase of acetylcholine in the hippocampus. Another suggested hypothesis is neuroplasticity. The hypothesis that acute fornix DBS could lead to long-term positive influence on memory has been tested here.  This was done through the enhancement of histological markers of neuro- and synaptic plasticity. Rats received DBS stimulation at 100 Hz, 100 μA and 100 μs pulse width for 4 h with electrodes placed bilaterally in the fornix. The water maze was performed five weeks thereafter.  Rats were sacrificed 6.5 weeks after stimulation.  BDNF, p-CREB, SV2 and synaptophysin immunohistochemistry was performed for their brains. No differences were found in the number of BDNF, p-CREB or SV2 positive cells for fornix DBS rats when compared to sham. Nonetheless, synaptophysin immunoreactive presynaptic boutons had a significant decrease in CA1 and CA3 of the hippocampus for DBS rats. Therefore, fornix DBS might induce long-term depression related mechanisms.

16:05 - 16:15 #15007 - O136 Nucleus basalis of Meynert neuronal activity in Parkinson’s disease.
O136 Nucleus basalis of Meynert neuronal activity in Parkinson’s disease.


Neuronal loss within the cholinergic nucleus basalis of Meynert (nbM) correlates with cognitive decline in dementing disorders such as Alzheimer’s disease (AD) and Parkinson’s disease.  In non-human primates, the nbM firing pattern (5-40 Hz) has also been correlated with learning and memory.  In this study, we performed microelectrode recordings of the bilateral globlus pallidus pars internus (GPi) and nucleus basalis of Meynert, and implanted deep brain stimulation (DBS) electrodes in Parkinson’s disease patients to treat motor symptoms and mild cognitive impairment, respectively. Here, we evaluate the neurophysiology correlates of the nbM in Parkinson's disease patients.


Three patients (3 male, age 68±3 years) with Parkinson’s disease and mild cognitive impairment underwent bilateral GPi and nbM DBS implantation.  Microelectrode recordings were performed through the GPi and nbM along a single trajectory. Firing rates and burst index were characterized for each neuronal population at rest and during an oddball cognitive task.


We characterized the firing rates of nbM cells (18±8 Hz), border cells (41±21 Hz), and GPi cells (70±46 Hz).  Firing rates of nbM cells were similar during an oddball task (13±10 Hz).  The burst index for nbM cells (1.36±0.14) was greater than border cells (1.11±0.05) but similar to GPi cells (1.24±0.14).  The nbM burst index was similar during the oddball task (1.34±0.15).


The trajectory through GPi and nbM has distinct neuronal firing patterns.  The profile of nbM activity is similar to that observed in non-human primates.  Further research is necessary to characterize the role of the nbM in cognition.

Darrin LEE (Toronto, CANADA), Robert DALLAPIAZZA, Luka MILOSEVIC, Philippe DE VLOO, William HUTCHISON, Suneil KALIA, Andres LOZANO
16:15 - 16:25 #16150 - O137 Prominent temporal coding of cognitive control in human prefrontal cortex.
O137 Prominent temporal coding of cognitive control in human prefrontal cortex.


In our daily lives, we are constantly faced with situations requiring us to make rapid yet accurate decisions. For example, when approaching a traffic signal that turns yellow, we have to attend to relevant information (speed, distance to the intersection, presence of police) and ignore the irrelevant information (radio, kids arguing in the back seat) in order to choose the optimal response (hit the accelerator or brake). These elements of “cognitive control” are critical components of normal cognition, and deficiencies in these processes underlie several neuropsychiatric disorders. Many studies have broadly attributed these cognitive control processes to prefrontal cortex, yet little is known about their neurophysiological basis.



We recorded single neuron firing rates and local field potentials (LFP) from the dorsal anterior cingulate cortex (dACC) and dorsolateral prefrontal cortex (dlPFC) in 19 patients undergoing neurosurgical procedures requiring intracranial electrodes, including epilepsy monitoring (N=12) and DBS (N=7). Subjects performed the multi-source interference task (MSIT), a Stroop-like task with 4 categories of decision conflict.



Neurons in both dACC (N=136) and dlPFC (N=367) demonstrated sparse firing rate coding of conflict level (dACC: 10.3%, dlPFC: 4.1%) (Figure 1). On the other hand, a majority of neurons encoded conflict level using a temporal code consisting of spike-field coupling to beta and theta oscillations (dACC: 49%, dlPFC: 52%). Spike-triggered LFP analysis indicated that firing rate encoding dACC neurons had the greatest effect on LFP in both dACC and dlPFC, suggesting that this small population of neurons may be entraining the LFP. Finally, spike-theta coherence in dlPFC predicted reaction time on a trial-to-trial basis, indicating a proximal relationship between neuronal activity and behavior (Figure 2).



The large majority of PFC neurons that do not demonstrate firing rate coding, previously assumed to be uninvolved bystanders, actually do participate in a temporal coding strategy that is both task- and behaviorally relevant. Temporal coding is emerging as a scheme used by several brain regions to facilitate communication and encode a wide dynamic range of information in a noise-robust manner. We propose that small populations of rate coding neurons (specialized “soloists”) entrain oscillatory potentials to recruit a larger population of temporal coding neurons (“choir”) that in turn stabilize and boost representations across the broader cognitive control network. These populations seem to work in concert to detect and communicate relevant information to optimize our decision-making capability.

Figure. Left: Sparse rate coding of decision-relevant variables in human dACC neurons. a, Microwire recording locations, with different colors per subject. b, Example dACC raster plot and firing rate over conflict conditions for a representative neuron that showed rate coding for decision conflict. c, Venn diagram for dACC neurons selective for specific elements of the task. Right: Reliability coding in human dlPFC neurons. a, Recording locations in the 9 patients from which dlPFC units were recorded b, Venn diagram c, Representative dlPFC neuron d, Representative coherogram for a single dlPFC neuron with no conflict and e, for both types of decision conflict. f, Difference between mean coherograms for trials with both types of decision conflict and trials without decision conflict. g, Mean difference coherograms averaged across all 367 dlPFC neurons. h, Mean difference coherogram illustrating significant difference in coherence across conflict conditions. i, Scatter plot and linear regression theta coherence predicting RT.

Sameer SHETH (Houston, USA), Guillermo HORGA, Garrett BANKS, Mark YATES, Yagna PATHAK, Guy MCKHANN, Elliot SMITH
16:25 - 16:35 #16189 - O138 Oscillatory coherent networks involving the internal Globus Pallidus and cortex.
O138 Oscillatory coherent networks involving the internal Globus Pallidus and cortex.


Internal Globus Pallidus (GPi) is part of the basal ganglia circuit and a common therapeutic target for the treatment of movement disorders. Deep Brain Stimulation (DBS) surgery affords an opportunity to study cortico-GPi coherent networks by simultaneous recording of GPi local field potential (LFP) and magnetoencephalography (MEG). The possibility to perform such recordings only exists in a few centres worldwide and to date only a single paper has been published (Neumann et al., Brain 2015). This study, done in dystonia patients, described three distinct coherent networks: pallido-temporal in the theta band (4-8 Hz), pallido-cerebellar – in the alpha band (7-14 Hz), and pallido-sensorimotor cortical – in the beta band (13-30 Hz). Here, we attempt to reproduce these results in a new and more heterogeneous patient group.




Nine patients were included in the analysis. Six of them had dystonic syndromes and three had Parkinson’s Disease but were recorded on dopaminergic medication when their symptoms were reduced.  Two of the dystonia patients had unilateral implantations resulting in 16 hemispheres with GPi recordings.

Resting MEG recordings were carried out using the Elekta Neuromag Vector View 306 channel system.


Coherence was computed in 0-45 Hz range between bipolar channels derived from adjacent GPi contacts and all magnetometer channels. Sensor-level statistical test was then performed in SPM12 ( to identify frequency bands with significant coherence. Source analysis for three fixed bands (theta 4-8 Hz, alpha 7-13 Hz, and beta 13-30 Hz) was performed with Dynamic Imaging of Coherent Sources (DICS) beamformer implemented in the DAiSS toolbox (  Images of coherence with left hemispheres were flipped in relation to mid-sagittal plane and all the images were included in a repeated measures ANOVA with factors frequency band and image type (original vs. surrogate). Regressors for subjects and hemispheres were included as confounding factors.


All results were significant with family-wise error correction (p<0.05) at the peak level. All significant coherence peaks were ipsilateral to the GPi electrode.




Two contacts of the electrodes were localized inside the posterior 1/3 GPi in all subjects. The distribution of the coherent frequencies at the sensor level showed two clear peaks: in the alpha-theta band peaking at 9Hz and in the beta band peaking at 25 Hz. In the theta band the main peak was found in the hippocampus with additional peak at the fusiform gyrus. In the alpha band there were two significant clusters: in the temporal lobe with peak in Brodmann area 22 (BA22) and in the brainstem. In the beta band one cluster was peaking in the premotor cortex (BA6) and an additional cluster was in the orbitofrontal cortex (BA11).




Our results are only partially consistent with the previous study. We also find coherence with inferior temporal lobe in the theta band and with motor areas in beta. However, there is no clear evidence for coherence with the cerebellum, and we find a new peak at BA11 in the beta band. The coherence topography in the alpha band is very similar to previously reported for the subthalamic nucleus (STN) which might suggest that STN and GPi are part of the same alpha network.

Chunyan CAO (Shanghai, CHINA), Dianyou LI, Yixin PAN, Shikun ZHAN, Vladimir LITVAK, Bomin SUN
16:35 - 16:40 #16223 - O139 A new type of thalamic potential related to somato-sensory system.
O139 A new type of thalamic potential related to somato-sensory system.


It has been postulated that high frequency oscillations (HFO) recorded at thalamus and elicited by somato-sensory evoked potentials (SSEP) are specific markers for the sensory ventro-caudal (Vc) nucleus. We have addressed this hypothesis by means of micro-electrode recordings (MER) during deep brain stimulation (DBS) of the centromedian nucleus (Ce) for refractory epilepsy.


SSEP elicited at contralateral median nerve were recorded in six patients anaesthetized (Propofol + remifentanil). Scalp (C3’/C4’-Fpz), cervical (C2-Fpz) and 4-leads MER recordings (bandwidth 2-5000 Hz, notch off) were averaged, starting at least 6 mm above the theoretical target at 1 mm steps. Reconstruction of MER position was done with the Schaltenbrand-Wharen atlas and using the final angles and coordinates of end DBS leads by means of MRi. Off-line analysis was performed using Matlab. Raw records were filtered at 2-200 Hz, to analyses local field potentials (LFP) and at 500-5000 Hz (Figure A), to analyses high frequency components (HFC). Spectral analysis (Fast Fourier Transform) was done onto both components (Figure 2B) and time-frequency analysis through Discrete Wavelet Transform (DWT) was applied onto HFC (Figure 2C).


LFP of two phases were observed in 60.7% recordings (N = 285), three phases in 29.1% records, four in 7.7% and finally, five phases 2.5% records. The first component appeared at 66% of records with a N1 phase (upward direction). The amplitude of LFP was 2.3 ± 0.1 µV ([1.1-2.7] range P25-P75), when the first component was N1 (upward), but when the it was P1 (downward), the amplitude was -2.1 ± 0.2 µV ([-2.3 - -1.0]), with a latency onset measured at peak of 19.43 ± 0.17 ms ([18.91 – 20.36]).

High Frequency Components. Can be divided into two components:

High Frequency Oscillations (HFO). Defined as those components whose frequencies are > 1 kHz. Are the second most frequent, (N = 224). It’s quite relevant to observe that HFO appears during long trajectories (up to 8 mm) and frequently in all the four electrodes. Spectra were slightly different from point to point. The amplitude was 3.0 ± 0.5 µV ([1.2 – 3.1]), with a latency onset of 14.29 ± 0.11 ms ([13.5 – 15.2]), a latency end of 25.37 ± 0.27 ms ([22.47 – 28.22]) and a frequency of 1471 ± 35 Hz ([1074 - 1850]).

Low Frequency Oscillations (LFO). We have defined the low frequency responses LFO as those potentials whose main frequency component does not exceed 1000 Hz, exhibiting amplitudes greater than HFO. These are the less frequent potentials, appearing in only 10 cases, usually in one point of recording, except in two trajectories were appeared in two points. The amplitude was 5.2 ± 1.8 µV ([3.7 – 7.9]), with a latency onset of 17.69 ± 0.47 ms ([16.11 – 18.89]), a latency end of 21.11 ± 0.27 ms ([19.55 – 22.95]) and a frequency of 848 ± 66 Hz ([700 - 865]).

Spectral analysis shows that components lower than 1 kHz are always present. However, it’s amplitude only is higher than peaks above 1 kHz when LFO are seen in the recording.

DWT for HFC recordings containing LFO and HFO show the presence of low frequency components (

Reconstruction of trajectories shows that sources of LFP and HFO are in several nuclei, including lamina medialis, ventral intermedium, centromedian and parafascicularis. However, the LFOs only appear in the sub-nuclei of V.c (Figure D).


We have shown that SSEP are composed by three components. Two of them (LFP and HFO) are highly scattered along several millimeters and in all the electrodes, with reconstructed sources located in several thalamic nuclei. However, the third component, LFO only appears in a highly localized manner and sources appear only in nuclei belonging to V.c. Therefore, we propose that LFO are the real physiological markers of somato-sensory synapsis at V.c. nucleus. This fact can shed light about the physiology of human sensory processing and help to improve accuracy of DBS surgery.

Jesus PASTOR (Madrid, SPAIN), Lorena VEGA-ZELAYA, Cristina TORRES, Marta NAVAS
16:40 - 16:45 #16268 - O140 Electrophysiological characterization of the centromedian nucleus in humans.
O140 Electrophysiological characterization of the centromedian nucleus in humans.

Introduction: The centromedian (Ce) nucleus has been used as target in deep brain stimulation (DBS) surgery. However, up to date, the electrophysiological properties of this nucleus have not been described. The aim of this study was to analyse the electrophysiological properties of the Ce nucleus in anesthetized patients.

Methods: By means of the projection of the theoretical reconstruction of the real trajectory through thalamus in the Schaltenbrand-Wharen atlas, we can identify the nucleus were electrode is located. We have analysed the neuronal recordings from 6 patients with refractory epilepsy undergoing bilateral DBS of the Ce nucleus, such as parvocellular (Ce.pc) as magnocellular ( sub-domains. All the extracellular action potentials (AP) have been sorted by Euclidian distances using amplitude (µV) and duration (ms) of the depolarizing and repolarizing phases (2 and 3, denoted as P2 and P3 respectively), e.g VP2, durP2, VP3, durP3. Then, we identify those pertaining to the same unit (clusterization) (Figure 1A). Finally, APs were merged to get the mean action potential (mAP). For all of these mAP, we computed the number of phases (2 or 3, if P1 is present), amplitude and duration, maximum and minimum values of first derivative (dVmax and dVmin, in mV/s) (Figure 1B) and the number of components of the repolarization phase (between maximum and minimum voltages) by means of the first derivative. Besides, we analysed the properties of raw discharge (e.g, mean frequency or firing rate, modified burst index -mBI-, pause index -PI- and pause ratio -PR). To conclude, we computed the density of cells (number of units at the raw record).

Results: A total of 362 mAP (146 for Ce.pc and 216 for from more than 2000 AP were obtained. We observed 3 phases in 66% y 75 % for Ce.pc y respectively therefore, 34% and 25% had only 2 phases, with a first negative phase (P1-) of shorter duration in most of cases, 72% in Ce.pc and 80% in (Figure 1C). The mean amplitude of P1- for Ce.pc was 19.8±1.4 µV and for 22.8±0.2 µV. The duration was 0.11 ±0.01 ms for both sub-domains. We found a first positive phase (P1+) in 28% and 20% for Ce.pc and respectively. The mean amplitude and duration was -17.8±1.6 µV and 0.15±0,01 ms for Ce.pc and -16.3±1.4 µV with 0.16±0.01 ms of duration for Regarding the second phase (P2-), we found a mean amplitude of 68.8±2.5 µV and 83.9±2.9 µV for Ce.pc and, respectively (p=0.008) with a duration of 0.37±0.01 ms for Ce.cp and 0.39±0.01 ms for (p<0.001). The mean amplitude of the third phase (P3+) for Ce.pc and was -38.3±1.6 µV and -45.8±1.6 µV (p=0.005) and the duration for Ce.pc 1.59±0.04 ms and for 1.75±0.03 (p=0.005). When we analyse the data of the total mAP, we found values of amplitude for the Ce.pc and of 107.1±3.9 µV and 129.8±4.5 µV (p=0.005). As for the duration, in the Ce.pc was 2.08±0.04 ms and in the was 2.26±0.03 ms (p=0.0002). The dVmax (mV/s) was for Ce.pc of 5.02±0.22 and of 6.08±0.24 for (p=0.023). Finally, the density of cells (units/raw) was 3.06±0.20 and 3.72±0.16 for Ce.pc and, respectively (p=0.015). No differences in mean firing rate, mBI, PI or PR were observed between both sub-domains.

Conclusions: There were significant differences between both sub-domains for amplitude and duration for P2 and P3 and for the total amplitude and duration of the mAP. Thus, the mAPs for both sub-domains exhibit different properties in morphology what make different and, therefore, easily recognizable by the analysis of their AP characteristics. The analysis of extracellular AP’s recorded during the surgery for DBS could help to identify the different nuclei, achieving a greater exactness in the placement of electrode. Therefore, we can increase the success in DBS improving the surgical outcome and decreasing the secondary effects. This technique must be especially relevant for thalamic surgery, where a great number of different nuclei are densely packed. The absence of difference in pattern discharge between Ce.pc and are not completely explained, but it is very important to remind that recordings were done under the effects of anaesthesia.

Lorena VEGA-ZELAYA (Madrid, SPAIN), Jesus PASTOR, Marta NAVAS, Cristina TORRES
16:45 - 16:50 #16282 - O141 Incidence of Seizures Induced by Intracranial Research Stimulation: a Multicenter Prospective Study in 770 Sessions Across 188 Patients.
O141 Incidence of Seizures Induced by Intracranial Research Stimulation: a Multicenter Prospective Study in 770 Sessions Across 188 Patients.

Introduction: Patients with epilepsy undergoing intracranial recordings provide an increasingly utilized opportunity to study human neurophysiology. Intracranial stimulation in these patients for research purposes can provide unique and valuable information, but ethical concerns demand a thorough appreciation of the associated risks. We measured the incidence of stimulation-associated seizures in a large multi-institutional prospective study using consistent stimulation parameters and seizure monitoring criteria.


Methods: 188 subjects who underwent intracranial epilepsy monitoring across 10 institutions participated in 770 stimulation sessions over 3.5 years. Seizures within 30 minutes of a stimulation session were considered potentially stimulation-related. For each observed seizure, we sought to determine whether it was likely related to stimulation or likely a naturally occurring seizure. To do so, we 1) analyzed the patient’s baseline seizure frequency to determine the statistical likelihood that the observed seizure was stimulation-related; 2) visually analyzed the temporal relationship between seizure onset and stimulation onset. Based on this scheme, we assigned each observed seizure to a category of definitely, possibly, unlikely, or definitely not related to stimulation.


Results: In total, 14 seizures occurred during or soon after a stimulation session (1.8% of sessions). Six seizures were similar to the patient’s typical seizures in terms of semiology, onset, and spread. All events were single seizures. The majority were simple partial (64%), or complex partial (29%); only one was generalized. No adverse events occurred, and length of stay in the monitoring unit was not affected. The mean amplitude of seizure-associated stimulation was 1.125mA (range 0.25-2mA), compared to a mean amplitude of 1.055mA (range 0.1-3.5mA) delivered in sessions without seizures. Using the categorization scheme, we found that 4 seizures (0.5%) were possibly stimulation-related, and 10 seizures (1.3%) were unlikely stimulation-related.


Conclusion: Seizures are a known possible risk of intracranial research involving brain stimulation. Using conservative criteria, we provide an upper limit of approximately 0.5-1.5% chance of stimulation-related seizure using our parameter range. The observed seizures did not add morbidity or affect the clinical course of any patient. These results will be important for understanding the feasibility and safety of intracranial stimulation for research purposes.

Hannah GOLDSTEIN, Elliot SMITH, Robert GROSS, Barbara JOBST, Bradley LEGA, Michael SPERLING, Greg WORRELL, Kareem ZAGHLOUL, Paul WANDA, Mike KAHANA, Dan RIZUTTO, Catherine SCHEVON, Guy MCKHANN, Sameer SHETH (Houston, USA)
16:50 - 16:55 #16300 - O142 Local field potentials in the thalamus and zona incerta during urinary functions.
O142 Local field potentials in the thalamus and zona incerta during urinary functions.

Background: Control of the lower urinary tract is complex and involves the integrated activity of distributed brain and spinal regions. However, the contribution of individual brain regions to bladder control is poorly understood. Implanted deep brain stimulation (DBS) electrodes enable the measurement of local field potential (LFP) signals from localised regions in the brain. Such recordings can provide insight into neurophysiological control of organ systems and pathophysiology of disease, and thus is relevant for the study of bladder control and dysfunction. Previous work has demonstrated measurable effects of DBS at the ventral intermediate nucleus of the thalamus (VIM) on urodynamic recordings during bladder filling (Kessler et al 2008), however, LFP analysis of VIM signals with regard to lower urinary tract behaviour has not been performed. 

Aims: To investigate neuronal oscillations in the VIM during imagined voiding, pelvic floor contraction and urinary voiding as well as their association with lower urinary tract symptoms. Signals from the zona incerta (ZI) were also recorded in patients with electrodes that spanned VIM/ZI, as a comparison nucleus.

Methods: 5 patients with VIM DBS were recruited; in 3 of these the electrodes also entered the ZI. 3 had essential tremor, 1 dystonic tremor and 1 Parkinson’s disease. LFPs were recorded during three experiments: imagined voiding, pelvic floor contraction/relaxation and urinary voiding. 

Bipolar channels were created by subtracting adjacent contacts for both the VIM and ZI. The signal was down-sampled to 1000 Hz in Spike2 and exported to MATLAB. The signal was low-pass filtered at 100 Hz, high-pass filtered at 2 Hz, and band-stop filtered at 50 Hz with a Butterworth filter. Power spectral density (PSD) analysis was performed on individual trials. These were then averaged per patient and then across patients. PSDs were also created for averages normalised by total power. Frequency bands were defined as: 2-4 Hz delta, 4-8 Hz theta, 8-12 Hz alpha,12-30 Hz beta, 30-90 gamma. Statistical comparison was done using Wilcoxon rank-sum test in individual patients and signed-rank test across patients.     

Linear regression in SPSS was used to assess the correlation between beta oscillatory power during voiding normalised to rest, and urinary symptoms assessed by the International Consultation on Incontinence Lower Urinary Tract Symptoms questionnaire was explored. 

Results: Significant frequency and frequency band changes were observed in the VIM and ZI during pelvic floor contraction/relaxation and imagined void at a single subject level, however, there was no significant change in LFP power during any experimental condition when signals were averaged across participants. 

Beta power during voiding normalised by resting beta power in the VIM, but not the ZI, significantly correlated with ICIQ scores for voiding (p=0.014), frequency (p=0.036) and incontinence (p=0.015) (Figure). 

Conclusions: There was no significant change in oscillatory power in the VIM or ZI during imagined void, pelvic floor contraction/relaxation or voiding. This implies that the VIM/ZI are not involved in normal voiding, however, our small sample size and heterogeneous population may have prevented us from detecting an effect. Significant bands in individual patients suggest hidden variables. Significant correlation between VIM beta power and lower urinary tract symptom severity in this patient group suggests that the beta signal may be of pathophysiological relevance for lower urinary tract symptoms.

Holly A ROY, Savva PRONIN (Edinburgh, UK), Yongzhi HUANG, Tipu Z AZIZ, James J FITZGERALD, Alex L GREEN
16:55 - 17:00 Discussion.

Friday 28 September

Added to your list of favorites
Deleted from your list of favorites
15:00 - 17:00

Parallel Session 14
Eating disorders & addiction

Moderators: Jane MORRIS (UK), Bomin SUN (director) (shanghai, CHINA), Pawel SOKAL (head of department) (Bydgoszcz, POLAND)
15:00 - 15:15 Basics of anorexia. Jane MORRIS (UK)
15:15 - 15:30 DBS for anorexia. Rebecca PARK (Associate Professor and Consultant Psychiatrist) (Oxford, UK)
15:30 - 15:45 Stereotactic ablation for anorexia. Roberto MARTINEZ-ALVAREZ (Neurosurgeon) (Madrid, SPAIN)
15:45 - 16:00 Basics of addiction. Christelle BAUNEZ (Director of Research) (Marseille, FRANCE)
16:00 - 16:15 DBS for addiction. Juergen VOGES (Head of the Department) (Magdeburg, GERMANY)
16:15 - 16:30 Surgical treatment for anorexia: DBS or ablation? Bomin SUN (director) (shanghai, CHINA)
16:30 - 16:40 #16406 - O143 Weight stabilization, based in normalization of the body mass index in 16 patients who underwent deep brain stimulation in the posterior hypothalamus for handling refractory extreme aggressiveness.
O143 Weight stabilization, based in normalization of the body mass index in 16 patients who underwent deep brain stimulation in the posterior hypothalamus for handling refractory extreme aggressiveness.

Introduction:  The management of patients with extreme refractory aggressiveness (ERA) has shown improvement with bilateral deep brain stimulation (DBS) in the posterior hypothalamus (PH).  In our experience, 16 patients who underwent surgery, not only improved their aggressive behaviour but also stabiliezed their weight, including both overweight and underweight.  Objective:  To report the clinical stabilization in the body mass index of 16 patients who underwent (DBS) in the PH due to ERA.  Methods:  16 patients operated between March 2013 and September 2017 with (DBS) in the (PH) due to (ERA) were followed-up with for a period of 8 to 62 months. Weight and height to calculate BIM were taken before surgery and betwen 6, 12, 24, 48 and 60 months after surgery. Other scales of quality of life, health status and aggressiveness were taken. a multidisciplinary group manages the patient before, during and after surgey. With Leksell frame implanted, Phililps 3 Teslas MRI is performed and the target is planned with Surgiplan software. 3387 Medtronic electrode is implanted bilaterally. Target:  X: 2 mm lateral to the lateral wall of the third ventricle, Y:  3 mm posterior with respect to the mid comissural point, Z:  2 mm above the upper edge of the red nucleus found with Fhc microrecording system.   Results:  The quality of life  EQ-5D-5L, the state of health ESH, the simplified scale of aggressiveness MOAS, as well the body mass index BIM, improved in the post surgical between 70-90%, 60-90%, 58-90%, 70-95% respectively.  Appetite disorders such as hyperphagia or not wanting to eat improved in the patients who presented it and some have become selective in the decision to consume specific foods that they did not have before.    Conclussions:  Bilateral (DBS) in the (PH) has shown significant improvement in aggressive behaviour and also in weight stabilization based on normalization of the body mass index in both overweight and underweight patients after surgery.  More studies and more cases should be carried out.  

Adriana Lucia LOPEZ RIOS (TORONTO, CANADA), Jonathan Ricardo DE LA CRUZ PABON, Alejandro ARISTIZABAL GAVIRIA, Luisa Fernanda AHUNCA VELASQUEZ, Gloria Elena RENDON PEREZ, Katherine Johanna NARANJO PEREZ, William Duncan HUTCHISON
16:40 - 16:50 #16358 - O144 Deep brain stimulation in the treatment of a patient with severe intractable anorexia nervosa – clinical experience and short-term follow up: Case report.
O144 Deep brain stimulation in the treatment of a patient with severe intractable anorexia nervosa – clinical experience and short-term follow up: Case report.

Objective: Anorexia nervosa is ranking within the highest mortality rates of psychiatric disease. Reports on safety of deep brain stimulation in patients with anorexia nervosa have documented clinical benefits of surgical therapy in a small series of patients. We have indicated to proceed with deep brain stimulation of the subcallosal cingulate as a rescue therapy in a patient with otherwise poor prognosis of life-time.


Material and Methods: The treatment option of deep brain stimulation was offered to a 20-year-old patient with treatment refractory anorexia nervosa who refused parenteral nutrition as well as nasogastric feeding. Baseline body mass index was 9.8. The patient was implanted with bilateral stimulation electrodes into the white matter of the subcallosal cingulate gyrus under general anesthesia. A rechargeable impulse generator was implanted in an infraclavicular subcutaneous pocket. A period of eight weeks of postoperative in-house surveillance is available for a short-term follow-up.


Results: No side effects, adverse events, or complications related to the surgical procedure occurred. Eight weeks after the operation the body mass index was stabilized at 12.1. Postoperative improvement in mood was reflected in a decrease in the Beck depression inventory scoring from 51 to 42. 


Conclusion: Deep brain stimulation seems to be a safe treatment option. A larger number of cases and long-term follow-up needs to be established to validate the long-term efficacy of the surgical treatment option for patients with severe, life-threatening anorexia nervosa.

Klaus NOVAK (Wien, AUSTRIA), Christoph KRAUS, Richard FREY
16:50 - 17:00 Discussion.