Background Radiofrequency (RF) thalamotomy remains an important treatment option for medication-refractory tremor. Modern MRI-based tractography enables individualized targeting of the dentato-rubro-thalamo-cortical pathway and may improve treatment consistency. However, data on outcomes of staged bilateral RF thalamotomy using modern imaging techniques remain limited, particularly regarding whether comparable tremor control can be achieved between the upper extremities. The aim of this study was to evaluate the efficacy and safety of tractography-guided RF thalamotomy and specifically to assess whether staged bilateral procedures result in symmetric tremor control. Methods We retrospectively analyzed 32 patients with medication-refractory tremor (Parkinson’s disease n=16; essential tremor n=16) who underwent tractography-guided RF thalamotomy between 2018 and 2024. Tremor severity was assessed using the Fahn–Tolosa–Marin Rating Scale (FTMRS). Clinical evaluations were performed preoperatively and at 6 weeks and 1 year postoperatively. Nine patients subsequently underwent staged bilateral procedures. The average time between unilateral and bilateral surgery was 14 months, with a minimum interval of 12 months. Results Following unilateral surgery, objective tremor in the treated upper extremity improved by 81% at the 6-week follow-up (p<0.001), with improvement in hands-on task performance of 63% (p<0.001). Response to treatment remained stable at 1 year. (Fig 1A). Among patients undergoing staged bilateral surgery, tremor on the second treated side improved by 56% (p=0.017), with a 30% improvement in task performance (p=0.020). At the one-year follow-up, tremor severity in bilaterally treated patients was symmetric between upper extremities, with no significant difference between sides (p=0.681). (Fig 1B). Adverse events were mainly mild and transient. After unilateral procedures, most adverse effects resolved during follow-up. After bilateral procedures, some patients had persistent mild symptoms such as balance difficulties or dysarthria, but severe complications were not observed. Conclusions Tractography-guided RF thalamotomy provides effective and durable tremor control. Staged bilateral procedures can achieve balanced and symmetric tremor reduction between upper extremities, suggesting that modern imaging-guided targeting enables consistent outcomes across hemispheres while maintaining an acceptable safety profile.

High-Intensity Focused Ultrasound (FUS) of the ventral intermediate nucleus (VIM) of the thalamus is an effective incisionless treatment for essential tremor (ET). However, clinical outcomes could remain variable, with potential side effects influenced by lesion placement. As the VIM can not be directly visualized on conventional MRI, current targeting relies on indirect methods including tractography. Advanced imaging has proposed direct targeting using a hypo-intense signal on White Matter Nulled (WMN) sequence, presumed to correspond to the target. Here, we aim to clarify the anatomical substrate of this hypo-signal and evaluate its relevance as an optimal target, alongside histological characterization. We retrospectively analyzed 100 ET patients treated with unilateral VIM-FUS. Clinical efficacy was quantified using the Clinical Rating Scale for Tremor (CRST) before, at 3 and 12 months after treatment Preoperative imaging included 3T MRI with WMN sequences and multishell diffusion acquisitions, from which a multimodal patient-based template was built for group analysis. Probabilistic tractography (MRtrix) was used to reconstruct the dentato-rubro-thalamic (DRT), pyramidal and lemniscal tracts. For each patient, the targeting strategies included the construction of the Guiot parallelogram, the registration of the deformable YeB atlas, the identification of the DRT barycenter and the WMN hypo-intensity visualization. Immediately after FUS, lesions were manually segmented on T2 MRI. Voxel-wise statistical analyses identified efficacy and adverse effect–related regions (“sweet and sour spots”). Ultra-high-field MRI was acquired from a fixed frozen postmortem human brain specimen. Photographs and immunohistochemistry on 40 µm sections enabled reconstruction of the cryoblock, co-registered to MRI. Normalization into the WMN template allowed anatomo-radiological comparison. Most patients were right-handed and underwent left-sided thalamotomy. Tremor reduction reached 90% at 1 month and 80% at 3 months. The WMN template revealed a consistent hypo-intense signal in the ventral lateral thalamus. The sweet spot overlapped this region. Histological examination confirmed myelinated DRT fibers entering the VIM at this location. These results indicate that the WMN hypo-intensity represents a key anatomical landmark corresponding to the terminal portion of the contralateral DRT rather than the VIM itself, suggesting a critical interface for therapeutic efficacy.

Background and Objectives: The target of thalamotomy for essential tremor (ET) is the Ventralis Intermedius (Vim) nucleus, a small area not clearly defined on Magnetic Resonance Imaging (MRI). We report our experience adopting high angular resolution diffusion imaging (HARDI) tractography for Vim targeting. Methods: Retrospective observational cohort study on 18 consecutive Gamma Knife Stereotactic (GKRS)-thalamotomies for ET. Dento-Rubro-Thalamo-Cortical-tract (DRTT), Cortico-spinal tract (CST) and Medial Lemniscus (ML) were reconstructed and used to adjust the coordinates-based target. Results: The median maximum dose to CST was 13.7 Gy (IQR 10.9-19.7 Gy) and the median CST-volume receiving > 20 Gy (V20) was zero (IQR 0-0.7 mm3). Conversely, the DRTT received a median maximum dose of 116 Gy (IQR 92.8-124.2 Gy) and the median V20 was 154 mm3 (IQR 63.5-190.2 mm3), whereas the median corresponding values for the ML were 31.5 Gy (IQR 25.7-53.2 Gy) and 16 mm3 (5.0-53.0 mm3), respectively. The maximum dose to the Vim was 126 Gy (IQR 126.2-127 Gy) and the median volume covered by 100 Gy was 25 mm3 (IQR 23-28-2 mm3), corresponding to the 14.5% of the whole volume (IQR 10.7-18.0%). The median Euclidean distance between the indirect (in-Vim) and adjusted Vim (ad-Vim) was 1 mm (IQR 0.7-1.9) and the final position of the ad-Vim was 1mm cranial compared to the in-Vim (p<0.001). Tremor improved in 15 of 18 thalamotomies (83.3%) and in all 11 cases with follow-up longer than 12 months, with a median improvement of 63.7% (IQR 57.4-72.3). The only prognostic factor was higher dose rate (p = 0.007), associated with faster response. No patients presented adverse radio-induced events during the follow-up. Conclusions: GKRS thalamotomy adopting DRTT projection, as landmark for the Vim in addition to standard coordinates, and maintaining the CST outside the 20 Gy isodose line, is effective and safe in ET.

Background: Current treatments for Alzheimer's Disease (AD) target downstream amyloid-β aggregation and plaque deposition in patients with MCI and mild AD. We evaluated single intracerebroventricular (ICV) injections of ex vivo-expanded autologous adipose-derived stem cells (ADSCs) prepared using a defined Wnt-activation protocol (RB-ADSC, the test product) in participants with mild or moderate AD. Methods: In a 3-plus-3 dose-escalation Phase I trial, six patients aged 45-80 years with mild or moderate AD (Functional Assessment Staging Tool (FAST) stages 4-5), underwent lipoaspiration, Ommaya reservoir placement, and an ICV injection of 2 or 5 million autologous Wnt-activated ADSCs. Safety was the primary endpoint; secondary endpoints included CSF biomarkers, amyloid PET centiloid scores, and ADAS-Cog-13 scores. Results: Adverse events (AEs) from liposuction and Ommaya reservoir implantation included transient mild bruising and discomfort. There were no grade 3 or 4 AEs or serious AEs associated with the RB-ADSC injections or treatment up to a 16 month follow-up. Favorable changes in CSF biomarkers, PET imaging, and cognitive scores were observed. From baseline to 12-week follow-up, the median CSF Aβ42 decreased 59.6%, 309.4 pg/ml (range: 276.1 – 370.4 pg/ml) to 125.0 pg/m (125.0 – 329.4 pg/ml); p-Tau decreased 52.2%, 61.9 pg/ml (46.9 – 152.8 pg/ml) to 29.8 pg/ml (13.2 – 66.6 pg/ml); total tau decreased 54.6%, 346.0 pg/ml (213.2 – 1190.6 pg/ml) to 157.2 pg/ml (50.0 to 351.6). CSF histone and nFL levels decreased a median of 84% and 13% respectively, and BDNF levels increased a median of 1.4%. Median amyloid PET scan centiloid score decreased 23.6%, 125.4 (range: 53.3 to 155.5) to 95.8 (range: 55.8 to 168.0); median ADAS-Cog-13 cognitive scores improved 34%, 58.5 (range: 40 to 69) to 38.5 (range: 20 to 69). Conclusions: ICV injections of ADSCs were safe and well-tolerated. Exploratory biomarker and cognitive findings support further evaluation in randomized controlled trials in AD and other neurodegenerative diseases. A Phase 2 trial has begun.

Background: Programming stimulation parameters is crucial for maximizing therapeutic benefit while minimizing adverse effects in patients with Parkinson’s disease (PD) undergoing deep brain stimulation (DBS). Beta-band power measured by local field potentials (LFPs) is associated with PD pathophysiology and has been proposed as a potential biomarker for optimizing stimulation parameters, particularly in the subthalamic nucleus DBS. However, its utility in globus pallidus internus (GPi) DBS remains unclear. Objective: To assess the utility of beta power in LFPs recorded from implanted deep brain electrodes during programming sessions after GPi-DBS implantation for PD. Methods: LFPs were recorded during follow-up visits in patients with PD who underwent GPi-DBS. Contacts were screened according to beta power, under the hypothesis that contacts showing a beta power peak may provide optimal clinical outcomes. The results were compared with those obtained by conventional contact screening based on neurological examination. In cases where the contacts suggested by beta power differed from those selected by conventional screening, potential additional benefits were evaluated using the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale Part III. Results: A total of 20 consecutive patients with PD who underwent GPi-DBS were included in the study. Of these, 55% were male; the mean ± standard deviation age at surgery was 71.6 ± 9.7 years, disease duration at surgery was 7.1 ± 2.2 years, and follow-up duration at the last programming session was 18.3 ± 8.5 months. A total of 40 electrodes were analyzed, and the optimal contacts identified by beta power were consistent with those selected by conventional contact screening in 72.5% of electrodes. Although additional benefits of directional stimulation guided by beta power spectra were unclear in most cases, three patients showed potential further improvement. Conclusion: Programming after GPi-DBS based on beta power in LFPs may help identify clinically useful contacts and could serve as a supportive tool for contact selection. However, its incremental benefit over conventional clinical screening appears limited. Further validation in larger cohorts is needed to clarify its role in optimizing GPi-DBS settings.

Objective: To delineate hemisphere-specific, local, and network-level mechanisms underlying stimulation-induced speech impairment after subthalamic deep brain stimulation (STN-DBS). Background: STN-DBS can improve but also worsen speech, suggesting that stimulation can modulate cortico–basal ganglia speech networks. However, the network mechanisms underlying these effects remain poorly understood. Methods: 25 patients with STN-DBS were assessed OFF medication during chronic stimulation and during a blinded randomized protocol in which stimulation amplitudes were systematically varied across 8 stimulation conditions above and below an individually defined speech-impairment threshold, yielding 190 right- and 189 left-hemispheric stimulation conditions. Speech was evaluated using phonation, diadochokinesis, and reading tasks. Acoustic features were combined into a Composite Dysarthria Index and compared with blinded patient self-ratings. Results: Speech modulation increased dose-dependently and was more pronounced with left-sided stimulation. Objective biomarkers correlated with blinded self-ratings (R=0.53, p<0.001). Under chronic stimulation, the left-hemispheric speech “sweet spot” localized to the dorsolateral sensorimotor STN at the motor–associative interface (R=0.71, p=0.02) and was associated with motor STN–globus pallidus externus connections (R=0.52, p=0.01). In contrast, the right-hemispheric sweet spot was located more ventroanteriorly within associative STN (R=0.50, p=0.02), but no speech-improving pathways survived validation. Speech-worsening regions extended beyond STN borders into lateral, dorsomedial, and ventromedial territories. In the left hemisphere, speech impairment was linked to corticobulbar projections, pallidothalamic and cerebellothalamic tracts, and hyperdirect connections to prefrontal and anterior cingulate regions. Right-hemispheric impairment was largely confined to motor effector circuitry. Conclusions: Stimulation-induced speech impairment reflects engagement of an extended, left-lateralized cortico-subcortical network, whereas right-hemispheric effects are largely restricted to motor pathways. These findings provide a mechanistic framework for hemisphere-specific, volume- and fiber-informed STN-DBS programming and highlight the potential of digital speech biomarkers to guide image-informed and adaptive neuromodulation strategies aimed at preserving communication.

Deep brain stimulation (DBS) is an effective treatment in multiple dystonia subtypes. DBS offers the unique opportunity to record local field potentials (LFPs) from subcortical brain structures. Low frequency oscillatory activity (below 12Hz) is raised in dystonia and may correlate with symptoms in certain dystonia subtypes. Pallidal activity may be asymmetrical in cervical dystonia (CD) and can be related to the direction of abnormal head postures. We used the Medtronic PerceptTM sensing capable implantable pulse generators (IPG) to investigate LFP characteristics in a mixed dystonia cohort who underwent bilateral globus pallidus internus (GPi) DBS. Methods This was a prospective study from July 2022 to August 2025 in a national DBS centre. Sensing capable IPGs were implanted. LFP recordings were obtained on post-operative day 1 to 7 prior to stimulation initiation. Recordings were at rest with eyes open in the survey or livestream mode. Results Twelve patients with dystonia (5 CD, 7 non-CD) were included. Eleven patients had primary dystonia, with one tardive dystonia secondary to neuroleptic use with a cranial/oropharyngeal phenotype. Delta (0 – 4Hz) and theta (4 – 8Hz) peaks were present in CD and non-CD. When combined, only delta peak oscillations were evident. Theta peaks were present in 4/5 CD patients (bilateral in 3 patients) and 4/7 non-CD (bilateral in 3 cases), total 13/24 GPi. This difference was not statistically significant. Mean theta power was significantly higher in CD than non-CD (p = 0.0358). However, neither mean or peak theta power correlated with symptom severity overall or in the subgroups. No correlations were observed between symptoms and other frequencies (delta, low frequency (4 – 12Hz), beta or gamma). No significant interhemispheric GPi power asymmetry was observed at a group level in CD or non-CD in either 4–12 Hz or 13–30 Hz bands. However individual patients with CD demonstrated variable asymmetry, with moderate effect sizes. In contrast, non-CD showed minimal asymmetry and these between-group differences were not statistically significant. Conclusion Low-frequency activity was present across dystonia subtypes but did not reliably correlate with symptoms. Individual CD cases may demonstrate asymmetry although this was not present at group level. These results highlight the limitations of low-frequency oscillatory activity as a consistent electrophysiological biomarker in a heterogeneous dystonia cohort.

Postmortem brain samples remain central for investigating the pathophysiology and possible new biomarkers for Parkinson’s disease (PD). These samples enable single-cell methods but undergo postmortem degradation with alterations in protein and gene expression. We have demonstrated that samples can also be collected with the DeepCell method as byproducts of deep brain stimulation (DBS) surgery, using the tissue attached to the non-permanent surgical instruments, for example guide tubes and microelectrodes (Kangas et al., 2022). However, the number of cells that can be collected is limited compared to postmortem samples. We compared these two methods to determine whether comparable cellular signatures can be obtained using DBS-derived samples. The DeepCell group included 21 patients with PD who underwent bilateral DBS of subthalamic nucleus (STN) between 2018 – 2021 in Oulu University Hospital. Single microelectrode recordings through guide tubes were performed during the surgery to confirm the electrode placement. The brain tissue attached to the guide tubes was extracted and rapidly frozen. Proteomic analysis was done using liquid chromatography–mass spectrometry. This dataset was next compared to a previously published dataset by Zhu et al. (2024) to analyze cellular signatures. The Zhu et al. study included six patients with late-stage PD and six age-matched controls. The authors collected postmortem prefrontal cortex samples and used single-cell transcriptomic and proteomic analysis. This dataset was reanalyzed based on the significance values used in our dataset. From the DeepCell and Zhu et al. datasets, 4773 and 3737 unique proteins were identified, respectively. There was a significant overlap between the two datasets. The Zhu et al. dataset contained 60 differentially expressed proteins between PD patients and healthy controls. The majority of the differentially expressed proteins from Zhu et al. dataset were also detected in the complete DeepCell proteomics dataset (55/60). Of these, 25 were differentially expressed based on clinical parameters in our dataset, and five had the same direction of dysregulation in both datasets. These five proteins by their gene identifiers were NONO, HEBP2, CDH13, FAM114A2, and SERPINB9. Notably, NONO protein aggregates have been identified from neurons of patients with PD (Belur et al., 2024). See Figure 1. The DBS-based approach identified the majority of both total unique proteins and differentially expressed proteins from the postmortem samples. There were several common elements with similar dysregulation in both datasets. These results suggest that the DeepCell method is a feasible alternative for collecting postmortem samples, with the advantage of avoiding postmortem degradation.

Memory modulation is increasingly viewed as an emergent property of interactions between deep nuclei and distributed cortical systems, yet causal evidence in humans remains limited. Here we tested how deep brain stimulation (DBS) of two basal ganglia targets, the substantia nigra (SN) and the subthalamic nucleus (STN), differentially modulated episodic memory in Parkinson’s disease. Thirty-two patients undergoing DBS implantation were randomized to have the deepest contact on the DBS electrodes positioned within SN or within the ventral STN, and were assessed 1 month after surgery under OFF-state, 10 Hz, and 130 Hz stimulation during functional MRI and an episodic memory retrieval task. Stimulation produced target and frequency dependent cognitive effects: 10 Hz SN-DBS increased overall recognition accuracy, whereas STN-DBS at both 10 Hz and 130 Hz reduced recognition accuracy. Across patients in the OFF-state, subcortical to cortical coupling involving thalamus and putamen tracked interindividual variation in global cognition. DBS further reshaped thalamocortical and medial temporal to prefrontal interactions at rest and during retrieval, and these network changes predicted individual differences in stimulation related memory effects. Connectome based fiber filtering linked accuracy gains to prefrontal projecting pathways and accuracy costs to sensorimotor projecting pathways. Together, these results identify deep nucleus-to-cortex control routes through which DBS can bias episodic retrieval and provide a network-grounded framework for cognitive modulation—one that could guide individualized stimulation strategies to improve cognition.

Background: Huntington's disease (HD) presents a unique clinical challenge with coexisting hyperkinetic and hypokinetic symptoms, yet the underlying neural oscillatory mechanisms remain poorly understood. Objectives: To characterize pathological pallidal neural activity in HD and identify biomarkers for therapeutic optimization. Methods: We investigated pallidal oscillatory patterns in fifteen HD patients undergoing deep brain stimulation, recording video-synchronized local field potentials during symptom fluctuations and comparing findings with Parkinson's disease and dystonia patients. Results: HD exhibited distinct pallidal oscillatory signatures that differed from PD and dystonia. Theta power (2-8Hz) increased during hyperkinetic states while high beta power (20-30Hz) elevated during hypokinetic states, both correlating significantly with clinical symptom severity. These patterns were not modulated by voluntary movement. Electrophysiological connectivity analysis integrated with neuroimaging analysis revealed that GPe-GPi theta coherence correlated with indirect pathway structural connectivity while pallidal high beta power associated with direct pathway functional connectivity, reflecting HD's dual circuit pathology. Spatial mapping localized theta oscillations to the posterior globus pallidus, with fibers projecting to motor cortical areas. Conclusion: We establish an electrophysiological framework explaining HD's complex symptomatology through dual oscillatory signatures. These findings provide circuit-specific biomarkers for disease monitoring and anatomical targets for optimizing deep brain stimulation in HD patients.

Introduction: Essential tremor (ET) is a prevalent movement disorder often refractory to medical therapy. While Deep Brain Stimulation is a standard of care, Magnetic Resonance-guided Laser Interstitial Thermal Therapy (MRgLITT) has emerged as a minimally invasive stereotactic alternative. It allows for real-time MRI thermometric monitoring, ensuring controlled lesioning of the ventral intermediate nucleus (VIM) of the thalamus without the need for a permanent implant. We aimed to evaluate the 12-month efficacy, safety, and cognitive impact of robot-assisted unilateral MRgLITT thalamotomy. Methods: Twenty-one patients with pharmacoresistant ET underwent unilateral MRgLITT thalamotomy. Stereotactic planning was performed using 3T MRI, targeting the VIM based on AC-PC coordinates and the Schaltenbrand-Wahren atlas. A surgical robot was used for trajectory guidance and laser probe insertion (Visualase system). Real-time thermometry ensured precise thermal dose delivery. Assessments at baseline, 3 months, and 12 months included the Fahn-Tolosa-Marin Tremor Rating Scale (FTM-TRS), Quality of Life in Essential Tremor Questionnaire (QUEST-SI), Patient Global Impression of Change (P-CGI), and cognitive testing (MMSE and MoCA). Linear mixed-effects models were used for statistical analysis. Results: Mean age was 67.2 +/- 12.8 years. At 12 months, the treated limb FTM-TRS score improved by 75.1% (14.0 +/- 3.1 at baseline vs 3.5 +/- 1.6 at M12; p < 0.001). All patients (100%) achieved a greater than 50% tremor reduction. QUEST-SI improved significantly from 41.2 +/- 15.7% to 18.0 +/- 14.6% (p < 0.001). Patient satisfaction (P-CGI) reached 82.0 +/- 20.7%. Adverse events at 1 months (71.4%) were primarily transient (proprioceptive impairment, ataxia), with only one patient (4.8%) exhibiting persistent gait ataxia at 12 months. Cognitive evaluation showed a stable MMSE (27.2 vs 26.5; p = 0.158), while a slight decrease in MoCA was noted (25.4 vs 23.9; p = 0.024) without functional impact. No hemorrhages, infections, or rehospitalizations occurred. Conclusion: Unilateral robot-assisted MRgLITT thalamotomy provides robust and sustained tremor relief with a favorable safety profile. Real-time thermometry offers a unique level of surgical control over the lesioning process. Its minimal invasiveness and lack of hardware make it a compelling option in the stereotactic management of essential tremor.

Introduction: Adaptive Deep Brain Stimulation (aDBS) shows efficacy in improving motor symptoms via real-time neural feedback. However, evidence regarding its long-term clinical sustainability remains sparse, particularly in real-world cohorts transitioning from conventional DBS (cDBS). This study prospectively evaluates 36-month outcomes and tolerability of aDBS in a cohort previously stabilized on cDBS. Methods: Eighteen patients with advanced Parkinson’s disease (PD), treated with cDBS for at least 6 months before switching to aDBS, were prospectively analyzed using Percept PC. Assessments including MDS-UPDRS (Parts I–IV), LEDD, and PDQ-39 Summary Index (SI) were performed at Pre-DBS, cDBS-baseline (optimized settings), and at 6, 12, 24, and 36 months post-aDBS initiation. Linear Mixed Models (LMM) estimated longitudinal trajectories using Pre-DBS and cDBS-baseline as references (95% CI provided for key estimates). Results: Long-term tolerability was high (36-month continuation: 83.3%, n=15/18); three discontinuations were due to patient preference, not device failure. In the Med-off state, Part III motor symptoms showed significant, sustained improvement vs Pre-DBS, peaking at 12 months (Estimate: -24.3 [95% CI: -31.9, -16.7], p<0.001) and remaining robust at 36 months (-14.6 [95% CI: -22.4, -6.9], p<0.001). Importantly, motor complications (Part IV) and quality of life (PDQ-39 SI) demonstrated significant and durable improvement throughout the 36-month follow-up, confirming sustained suppression of motor fluctuations and dyskinesia. In contrast, ADL (Part II) significantly worsened at 36 months compared to the cDBS-baseline (Estimate: +5.0 [95% CI: 1.8, 8.3], p=0.004), despite remaining improved relative to Pre-DBS, reflecting the impact of disease progression on ADL. Non-motor symptoms (Part I) showed no significant improvement at 36 months compared to Pre-DBS. LEDD was significantly reduced and maintained throughout the follow-up period. Conclusions: aDBS is a well-tolerated therapeutic option providing durable benefits in motor symptom relief, suppression of motor fluctuations, and quality of life over 36 months. However, the decline in ADL (Part II) at 36 months relative to the cDBS-baseline underscores the limitations of the therapy in halting long-term disease evolution. While aDBS is highly reliable for motor and QOL management, clinical expectations regarding ADL outcomes must be grounded in the context of eventual disease progression.

Background: Subthalamic nucleus deep brain stimulation (STN‑DBS) is effective for Parkinson’s disease, but its therapeutic window may be restricted by current spread to the corticospinal tract (CST). Imaging‑guided workflows and algorithm‑based programming have been developed to refine contact selection, improve field shaping, and reduce CST recruitment. This study evaluated the concordance among clinical, imaging‑guided, and algorithm‑based programming, comparing motor side‑effect thresholds, therapeutic window, volume of tissue activated (VTA), VTA–CST interaction, and clinical outcome. Methods: Twenty patients (40 STN sides) implanted with the Vercise Genus™ system underwent three programming conditions: clinically selected optimal contacts (P1), imaging‑guided programming with Brainlab Elements GuideXT™ (P2), and algorithm‑based programming using the Illumina™ 3D algorithm (P3). Preoperative 3T MRI and postoperative CT were fused for lead localization and tractography. CST reconstruction used probabilistic DTI with ROIs in M1, internal capsule, and brainstem. VTA was modeled at the motor side‑effect threshold. Extracted metrics included motor thresholds, safety margin, VTA volume, VTA–CST overlap (mm³ and %), tangentiality, and overlap category. Each program was applied for one month in randomized order with optimized stimulation intensity and unchanged medication. Standardized clinical assessments were performed at each month’s end. Results: Active stimulation contact levels did not differ significantly across programs. Complete concordance (P1=P2=P3) occurred in over half of cases, with strongest agreement between P2 and P3. Motor thresholds were comparable. VTA volumes were smaller in P2 and P3 than in P1 (p<0.0001). VTA–CST overlap was significantly reduced in imaging‑guided and algorithm‑based programs in both absolute and percentage measures (p<0.0001). Tangential field orientation increased from 2.8% (P1) to 25.0% (P2) and 61.1% (P3). Despite similar VTA volumes in P2 and P3, P3 produced the lowest CST engagement. Conclusions: Imaging‑guided and algorithm‑based programming reduced CST involvement and promoted tangential, fiber‑aligned field geometries. Algorithm‑based programming showed the most consistent CST‑sparing profile, supporting its value in improving STN‑DBS tolerability. Further analyses of clinical benefit and longitudinal outcomes are required.

Background: Postoperative weight gain is a recognised complication of deep brain stimulation, but whether ablative surgery produces target-dependent weight effects remains unknown. We investigated whether pallidothalamic tractotomy (PTT) causes greater weight gain than pallidotomy and explored spatial associations. Methods: We retrospectively analysed 79 patients with focal, segmental or generalized dystonia who underwent unilateral PTT (n = 41) or GPi lesioning (n = 38). Multivariable linear regression quantified the effect of target on weight and BMI changes, adjusting for age, sex, side, preoperative BMI, follow-up duration, and BFMDRS improvement. In a subset with postoperative MRI, lesion overlap (N) maps and voxel-wise mean BMI-change maps were generated to explore anatomical associations. Results: PTT resulted in greater postoperative weight gain than GPi lesioning (+6.7 ± 7.5 kg vs +1.1 ± 4.7 kg), with ≥10 kg gain in 29.3% versus 5.3% of patients. After covariate adjustment, PTT remained independently associated with greater weight gain (adjusted difference +5.18 kg; 95% CI +2.13 to +8.23) and BMI increase (+1.92 kg/m ; 95% CI +0.85 to +2.99). Lesion mapping localised PTT lesions to Forel’s field H1. Weight-gain hotspots mapped medial to the inferior border of the subthalamic nucleus, whereas GPi maps showed no consistent intranuclear hotspot. Conclusions: Pallidothalamic tractotomy was associated with greater postoperative weight gain than pallidotomy, independent of covariates. These findings suggest that lesion target influences postoperative body-weight regulation, highlighting the potential risk of lesions extending into the caudal subthalamic region.

Objectives: Magnetic resonance–guided focused ultrasound (MRgFUS) has become an established minimally invasive treatment for essential tremor (ET) since the landmark randomized controlled trial in 2016. However, tremor recurrence remains common, and its mechanism is not fully understood. The dentato-rubro-thalamic tract (DRTT), a key component of the cerebello-thalamo-cortical network, can now be evaluated in detail using tractography, including both decussating (dDRTT) and nondecussating (ndDRTT) components. This study aimed to determine whether overlap between the MRgFUS lesion and DRTT components is associated with clinical outcomes in ET. Methods: We retrospectively analyzed patients with ET who underwent unilateral MRgFUS Vim thalamotomy between 2019 and 2025. Patients with at least one available pre- or postoperative DTI study were included. Clinical severity was assessed using the Clinical Rating Scale for Tremor (CRST), including separate analyses of parts A and B. Tractography was performed using BrainLab Elements 2.0. Regions of interest included the dentate nucleus, red nucleus, Vim, and primary motor cortex. Parameters were set at a minimum FA of 0.20, minimum fiber length of 80 mm, and maximum angulation of 20°. Overlap between the postoperative T2 lesion and dDRTT, ndDRTT, and Vim was calculated. Correlations between overlap measures and CRST changes were analyzed using Spearman’s test. Results: A total of 62 patients were included. The mean lesion volume was 0.12 ± 0.05 cm³. Improvement in total CRST score at 2 weeks was significantly correlated with overlap between the Vim and postoperative dDRTT (p = 0.0467, r = 0.3165). Improvement in CRST part B was significantly correlated with overlap between the Vim and ndDRTT (p = 0.0441, r = 0.3200). Lesion volume was associated with tremor recurrence in CRST part A between 2 weeks and 6 months (p = 0.025, r = −0.557). Conclusion: Greater overlap between the MRgFUS lesion and DRTT components was associated with early tremor improvement. Lesion volume was related to recurrence, suggesting that both tract targeting and lesion size influence outcomes after MRgFUS.

Background: Adaptive deep brain stimulation (aDBS), which automatically adjusts stimulation amplitude based on local field potentials (LFPs), has the potential to provide individualized therapy according to symptom fluctuations. However, its objective superiority over conventional DBS (cDBS) remains unclear. While prior comparisons at our institution using clinical scales showed no significant differences, continuous objective assessment of motor activity has not been fully explored. Methods: Seven patients with Parkinson’s disease implanted with a Percept™ PC or RC system were evaluated at 3 months postoperatively. Using optimized cDBS settings as a reference, patients underwent both aDBS and cDBS for 24 hours each during the same hospitalization period. Motor activity was continuously recorded using a wrist-worn accelerometer (ActiGraph) with 60-second epochs. Within-subject comparisons were performed using the Wilcoxon signed-rank test. Results: No significant differences were observed in overall 24-hour motor activity between conditions. However, time-of-day analysis revealed significantly higher energy expenditure (kcal) during the morning period (06:00–12:00) under aDBS (p = 0.031), with a trend toward increased vector magnitude (Vm) (p = 0.063). Mean stimulation amplitude was significantly lower in aDBS compared to cDBS. Conclusion: While overall motor activity was comparable between aDBS and cDBS, aDBS demonstrated enhanced morning activity, suggesting a potential role in modulating diurnal motor patterns. These findings highlight the value of continuous wearable monitoring for detecting subtle temporal effects of aDBS. Future studies incorporating long-term home-based monitoring and correlation with LFP dynamics are warranted.

Introduction MR-guided transcranial focused ultrasound therapy(MRgFUS) is an incisionless treatment that increases tissue temperature by converging multiple ultrasound beams and corrects pathological neural networks through thermal coagulation. Differences between T2-weighted imaging(T2WI) and susceptibility-weighted imaging(SWI) become evident approximately 3 months after FUS. While lesions almost disappear on T2WI, they remain detectable for a longer period on SWI. While the associations between postoperative T2WI lesion size and clinical outcomes has been reported, data regarding SWI lesions remain limited. This retrospective study evaluated thalamic lesions after MRgFUS using T2WI and SWI and examined their associations with long-term clinical outcomes. In addition, microelectrode recordings(MER) were obtained in one recurrent case and compared with MRI findings. Methods Patients with medication-refractory Essential tremor(ET) or tremor dominant Parkinson’s disease(TDPD) who underwent MRgFUS thalamotomy and received SWI at 6 or 12 months postoperatively were included. Lesion size was measured on immediate postoperative T2WI and long-term SWI. Clinical outcomes were assessed using the Clinical Rating Scale for Tremor(CRST) for ET and Unified Parkinson's Disease Rating Scale(UPDRS) items 16, 20, and 21 for PDT. Multiple regression analysis was performed with long-term clinical outcome as the dependent variable, including lesion size and skull density ratio as explanatory variables. Results Seventy patients were included (54 ET and 16 TDPD). In ET, the CRST score improved from 26 ± 12 to 10 ± 7 and worsened to 12 ± 9 at long-term follow-up. In TDPD, the UPDRS score improved from 8 ± 3 to 1 ± 2 and increased to 3 ± 2 over time. In multiple regression analysis, long-term SWI lesion size showed a trend toward a positive association with long-term clinical outcomes compared with immediate postoperative T2WI lesion size; however, neither association reached statistical significance. One patient in his 70s with ET underwent deep brain stimulation 28 months after MRgFUS. MRI showed a smaller T2WI lesion within a larger SWI lesion. MER demonstrated typical Vim neuronal activity within the SWI lesion, whereas the T2WI lesion corresponded to a region with markedly reduced neuronal activity. Conclusion Long-term clinical outcomes after MRgFUS thalamotomy may be more closely associated with lesions depicted on long-term SWI than with immediate postoperative T2WI lesions.

Introduction: Deep brain stimulation (DBS) is an established therapeutic option for severe, treatment-refractory dystonia and may be considered in cases of rapidly worsening dystonia or status dystonicus despite optimized medical management, particularly in certain monogenic forms (DYT-TOR1A, GNAO1, KMT2B). We report our pediatric neurosurgery department’s experience with emergency DBS in patients presenting with pre-status or status dystonicus and describe their medium- to long-term outcomes. Methods: We performed a single-center retrospective review of all patients who underwent emergency primary DBS implantation between 2015 and 2025. Clinical, genetic, perioperative and follow-up data were extracted from medical records. Results: Nine patients aged 5–20 years underwent urgent bilateral pallidal DBS for pre-status or status dystonicus. Eight had monogenic dystonia (GNAO1, DYT-TOR1A, NUP54, ADCY5, ZBTB11); in three, the genetic diagnosis was established after DBS implantation. One patient had dystonia secondary to hypoxic-ischemic injury. At presentation, three patients had grade 5, four grade 4, and two grade 3 on the Dystonia Severity Action Plan scale. The mean pre-implantation hospital stay was 20 days. Stimulation was initiated within 24 hours in all cases, with progressive amplitude increases to a median of 1.7 mA at discharge. All patients showed clinical improvement sufficient for ICU discharge after a mean of 9 days (range 2–28). Postoperative hospital stay was variable, (mean 3 days, SD 32). Antidystonic medications decreased from a mean of 5 to 4 during hospitalization. Two postoperative complications occurred (one transient respiratory distress episode and one surgical site infection). After a mean follow-up of 4.2 years, four patients experienced recurrent status dystonicus: two due to hardware dysfunction (battery depletion), resolving after stimulator replacement; one following explantation for infection; and one patient with hypoxic-ischemic injury who relapsed despite a functioning device during medication tapering. Conclusion: Emergency DBS appears as a feasible and effective rescue option for status dystonicus or rapidly worsening dystonia unresponsive to medical management. It enabled resolution of the acute episode in all patients and contributed to sustained symptom control in most cases. Given the risk of relapse, particularly in the setting of hardware failure or medication adjustments, close long-term monitoring is essential.

Background: High-frequency focused ultrasound (HFUS) thalamotomy is an established, incisionless treatment for medication-refractory essential tremor (ET) that provides robust tremor reduction. However, transient gait and balance disturbances are commonly observed following the procedure. Despite this, there remains limited quantitative evidence characterizing peri-procedural balance changes and the trajectory of recovery. Objective: To prospectively evaluate balance changes following HFUS thalamotomy using the Mini-Balance Evaluation Systems Test (Mini-BESTest) and to define the temporal profile of recovery. Methods: We conducted a prospective study of 100 patients with medication-refractory ET undergoing unilateral HFUS thalamotomy. Balance was assessed using the Mini-BESTest (maximum score: 28) at four time points: pre-procedure (baseline), postoperative day 1, 2 weeks, and 3 months. Tremor severity, including postural and intention components, was assessed clinically at each interval. Changes in Mini-BESTest scores were analyzed descriptively to evaluate patterns of balance impairment and recovery over time. Results: Ninety percent of patients demonstrated significant improvement in postural and intention tremor immediately following HFUS, with sustained benefit at 3 months. Mini-BESTest scores declined on postoperative day 1, reflecting acute, transient balance impairment. By 2 weeks, mean scores returned to baseline levels across the cohort, indicating recovery of balance function. At 3 months, balance remained stable with no evidence of persistent deficits. No patients experienced clinically meaningful long-term balance deterioration. Conclusions: Unilateral HFUS thalamotomy for essential tremor is associated with a predictable, transient decline in balance immediately post-procedure, with recovery to baseline typically occurring within 2 weeks. These findings provide objective data to inform patient counseling regarding expected postoperative balance changes. Importantly, early balance disturbances do not appear to compromise long-term functional outcomes, as durable tremor control is achieved without persistent balance impairment.

Background: Head tremor may occur in isolation or within essential tremor spectrum, dystonic tremor, Holmes tremor, and secondary tremor syndromes. Dedicated DBS series focused specifically on head tremor remain limited. Methods: We retrospectively reviewed 32 patients treated with double target VIM–PSA DBS between 2019 and 2025 for clinically relevant head tremor. Head tremor severity was assessed with the TRS head subscore preoperatively and at follow-up under chronic stimulation, with paired postoperative DBS off/on analyses. Postoperative lead reconstruction, VTA modeling, sweetspot mapping, and connectomic fiber engagement analyses were performed in Lead-DBS. Results: Mean head TRS improved from 2.41±0.95 preoperatively to 0.47±0.72 at follow-up under chronic stimulation. In paired postoperative assessments, mean head TRS decreased from 2.50±0.92 in DBS off to 0.47±0.72 in DBS on, with complete suppression in 21/32 patients (65.6%). Benefit was observed across tremor types and clinical presentations and was most favorable in essential tremor spectrum, dystonic tremor, isolated head tremor, and no-no directional pattern. TRS-based sweetspot mapping identified a benefit-related sweetspot centered below VIM and extending to PSA within the VIM–PSA corridor, while connectomic fiber recruitment shifted systematically with stimulation position across the corridor. Conclusions: Double target VIM–PSA DBS was highly effective for head tremor in a clinically heterogeneous cohort. Clinical and connectomic findings support a corridor-based interpretation in which benefit is best explained by programmable access to both VIM and PSA, including simultaneous VIM and PSA stimulation when required, rather than superiority of either compartment in isolation.

Introduction Deep brain stimulation (DBS) significantly improves the motor symptoms of Parkinson’s disease, as well as dyskinesias induced by dopaminergic treatment. The two main surgical targets are the internal globus pallidus (GPi) and the subthalamic nucleus (STN). The advent of directional electrodes has improved clinical outcomes in patients treated with STN DBS, notably by reducing side effects related to current spread to adjacent structures. However, no clear superiority of these directional electrodes has been demonstrated to date for GPi DBS. Objectives To evaluate, in a randomized crossover pilot study, the chronic clinical effect of different stimulation modes (directional versus omnidirectional, mono or bipolar), following implantation of Boston Scientific Cartesia™ electrodes in Parkinsonian patients eligible for GPi DBS. Methods We included 9 patients with idiopathic Parkinson’s disease meeting criteria for DBS surgery. At one month postoperatively, contact-by-contact testing was performed to determine efficacy and side-effect thresholds; patients were then randomized to either the ring stimulation group or the directional stimulation group. Six weeks later, the type of stimulation was switched, and six weeks thereafter, stimulation was set to bipolar mode. At the end of the testing period, the best stimulation configuration was selected and reassessed at one year postoperatively. The primary endpoint was improvement in motor symptoms and dyskinesias assessed using Part III and IV of the UPDRS scale. Results The mean UPDRS III (ON MED) was 38.25 (±15.47) in the overall population at the initial screening visit. The mean UPDRS III (ON MED) at the 3-month visit (V2) was 16.44 (±9.33) in the overall population. No significant difference in UPDRS III was observed between the two groups at this time point. The mean initial UPDRS IV (ON MED) was 10.11 (V0) versus 2.14 at (V2). An average reduction of 90% in dyskinesias was measured at 1-year. Statistical analysis showed no significant difference between directional and ring stimulation at 3 months and at 12 months. At the end of the testing period, patient choice of stimulation mode did not show a significant difference. Conclusion There does not appear to be superiority of directional GPi stimulation in patients with motor symptoms of Parkinson’s disease. One possible interpretation relates to target size, as the GPi is larger than the STN, resulting in less current spread to adjacent structures.

Introduction Focused ultrasound (FUS) thalamotomy is an effective, minimally invasive surgery for essential tremor (ET). However, this procedure causes chronic sensory deficits in approximately 14% of patients, likely due to involvement of posterior sensory thalamic nuclei. This study compared thalamic targets predicted by an artificial intelligence (AI) to the locations chosen by neurosurgeons, and whether AI-lesion vector offsets were associated with persistent sensory deficits. Methods Seventeen ET patients who successfully underwent FUS thalamotomy but developed chronic sensory deficits (median follow-up 11.7mo; IQR 10.1-14.7) were matched to eighteen similar patients without deficits (median follow-up 11.4mo; IQR 10.2-13.2). All patients were treated at a single institution between the years 2011 and 2025. MRI-derived coordinates of lesion locations were retrospectively compared to AI-predicted locations (RebrAIn OptimMRI software). Differences in lesion characteristics and tremor outcomes were analyzed using appropriate statistical and regression methods. Results AI-predicted lesions were significantly more anterior in sensory deficit patients (mean offset 1.10 ± 0.81 mm) compared to the control group (0.33 ± 0.56 mm, p=0.003). Side-effect status was independently associated with difference in AP AI–lesion offset (β=0.75mm, p=0.005), whereas lesion volume was not (β=0.00mm, p=0.85). The sensory deficit group also demonstrated more lateral lesion locations (14.7 ± 0.9mm vs 13.9 ± 0.6mm, p=0.004) and AI-predicted targets (15.1 ± 0.9mm vs 14.4 ± 0.8mm, p=0.02). However, mediolateral AI-lesion offset did not differ between groups. There were no significant differences in lesion characteristics between groups along the superior–inferior axis. Tremor was less improved in patients with sensory deficits (56.7 ± 44.8%) than in those without (86.0 ± 25.4%, p=0.027), with total recurrence also being more frequent (6/17 vs 1/18, p=0.041). Side-effect status was independently associated with difference in tremor outcome (β = −28.171%, p = 0.045), whereas lesion volume was not (β = −0.001%, p = 0.988). Conclusions AI correctly predicted more anterior targets in patients with chronic sensory deficits and worse tremor outcomes following FUS thalamotomy. These findings suggest that AI-assisted targeting could improve FUS thalamotomy accuracy and reduce sensory complications caused by suboptimal lesion placement.