Enhancing Asleep Deep Brain Stimulation Targeting for Essential Tremor Using Machine Learning: The OPTIVIM Phase 2 Study

Abstract

BACKGROUND AND OBJECTIVES:

Deep brain stimulation (DBS) is an effective treatment of essential tremor, but the optimal target and how to reach it with the best accuracy remain controversial. The OptimDBS algorithm is based on Kernel Ridge Regression and was trained on a database of patients who underwent DBS with optimal postoperative outcomes. Using 3-dimensional T1-weighted MRI as the only input, it calculates the stereotactic coordinates of an effective DBS target for treating tremor. The aim of the study was to evaluate the efficacy and safety of OptimDBS-guided asleep DBS without peroperative clinical and electrophysiological controls in an independent cohort of patients.

METHODS:

The OPTIVIM study was a prospective, single-arm, multicenter Phase 2 trial. The primary outcome was tremor reduction and improvement in activities of daily living as assessed by the Fahn-Tolosa-Marin scale 3 months after surgery. Secondary outcomes included reduction in tremor amplitude on accelerometry, quality of life (modified Parkinson’s Disease Questionnaire-39), ataxia (Scale for the Assessment and Rating of Ataxia), adverse events, and anatomic location of active contacts and the volume of tissue activated, all measured 3 months after surgery.

RESULTS:

Twenty-two patients from 2 centers were enrolled. Fahn-Tolosa-Marin scores improved by 61.3% (95% CI: 53.7%-68.9%). Accelerometry (mean ± SD) showed a reduction in postural tremor amplitude of 81% ± 56% on the right side and 84% ± 35% on the left side. Modified Parkinson’s Disease Questionnaire-39 scores (median [Q1-Q3]) improved by 55% [24.3%-77.8%]. Scale for the Assessment and Rating of Ataxia scores remained stable (median [Q1-Q3], preoperative vs postoperative: 4 [3-6] vs 3 [2-6]). There were no serious adverse events. Active contacts were located in the posterior subthalamic area (59%) or the ventral-intermediate nucleus of the thalamus (20%), with 100% of volume of tissue activated overlapping the ventral-intermediate nucleus of the thalamus or posterior subthalamic area.

CONCLUSION:

The results of OptimDBS-guided asleep DBS seem to be comparable with those reported in the literature for awake and asleep DBS and ablative techniques. Long-term evaluations with larger cohorts are needed to confirm these results.

ABBREVIATIONS:

ADL

activities of daily living

AEs

adverse events

CoP

center of pressure

ET

essential tremor

FTM

Fahn-Tolosa-Marin

HIPS-THOMAS

Histogram-based Polynomial Synthesis-Thalamus Optimized Multi-Atlas Segmentation

mPDQ

modified Parkinson's Disease Questionnaire

PSA

posterior subthalamic area

SARA

Scale for Assessment and Rating of Ataxia

VIM

ventral-intermediate nucleus of the thalamus

VTA

volume of tissue activated.

Essential tremor (ET) is a common movement disorder¹ that can significantly impair quality of life in advanced stages. Deep brain stimulation (DBS) targeting the dentatorubrothalamocortical pathways, particularly at the ventral intermediate nucleus and the posterior subthalamic area, is considered the most effective intervention.¹ Awake DBS, guided by clinical testing and microelectrode recordings, is still considered the gold standard. Yet, variability across indirect targeting methods—due to differences in anatomic atlases,^2,3 statistical coordinates,^4,5 and tractography techniques⁶—has led to inconsistent results. However, some studies suggest that asleep procedures without intraoperative testing may yield comparable outcomes with those of awake surgery.^7-9 In this context, we introduced the concept of a clinically defined target, derived from outcomes in successfully treated patients. We developed the OptimDBS algorithm (RebrAIn©) using a machine learning approach trained on a cohort of patients who experienced excellent outcomes after DBS for ET.^10,11 For each hemisphere, the algorithm learned the spatial relationship between the active contact and 18 landmarks on T1-weighted MRI (corresponding to the edges of the thalamus, the putamen, the habenular commissure, and the mammillothalamic tract). In a new patient, OptimDBS detects the same landmarks and, using the learned model, predicts the most likely location of the active contact—thus providing a personalized stimulation target without the need for advanced imaging. This makes the software accessible to a broad range of surgical centers equipped with conventional MRI and stereotactic planning tools. After previous internal and external validations,¹¹ the OptimDBS algorithm is now being prospectively evaluated to assess its efficacy and safety.

METHODS

Study Design

OPTIVIM was a single-arm, prospective, unblinded, bicentric, phase 2 clinical trial with an intention-to-treat analysis. The study report is described in accordance with consolidated standards of reporting trials in artificial intelligence extension.¹²

Patients

The inclusion criteria were as follows: advanced, disabling, and pharmaco-resistant ET; age between 18 and 75 years; normal cerebral MRI; a Mattis Dementia Rating Scale score¹³ ≥130; and affiliation with the national social security system. Exclusion criteria included major depressive syndrome (Beck Scale¹⁴ score >20), any surgical or anesthetic contraindication, MRI findings of significant cerebral atrophy, or extensive fluid-attenuated inversion recovery hypersignals. Pregnant and breastfeeding women were excluded.

Procedure

Target identification was performed on 3-dimensional (3D) T1-weighted MRI by the OptimDBS algorithm (see next paragraph). The MRI scan could be conducted either in the days preceding surgery (preoperative MRI) or simultaneously with frame placement (stereotactic MRI). Surgery was performed under general anesthesia. In this protocol, the neurosurgeon was required to implant the electrode directly at the target predicted by the algorithm without any intraoperative adjustment. No intraoperative electrophysiological monitoring or clinical assessment was performed. The procedure was completed in a single step, involving bilateral lead insertion and stimulator implantation, following a previously described protocol for Parkinson disease.¹⁵ The target was planned to be in the middle of the 4 contacts of the lead (Medtronic™ 3389). DBS parameters were adjusted according to each center's standard practices. Patients could return for additional stimulation adjustments between hospital discharge and the 3-month follow-up if needed. All medications were permitted throughout the trial.

OptimDBS Algorithm

OptimDBS is embedded in a dedicated software platform (OptimDBS; RebrAIn©) developed for this study. It uses 3D T1-WI as input to identify 18 anatomic landmarks per hemisphere (a full list is provided in the Supplementary Material, http://links.lww.com/NEU/F126). The algorithm uses kernel ridge regression to map the 3D configuration of these landmarks to a single set of stereotactic coordinates in anterior commissure–posterior commissure corresponding to the predicted target (the mathematical method is detailed in the Supplementary Material, http://links.lww.com/NEU/F126). This algorithm was trained on a data set of patients who had undergone successful DBS for ET, see11 for details. For each new patient, the software labels the target on the native MRI with a 3D marker.

Judgement Criteria

All end points were assessed preoperatively and at 3 months postoperatively. The primary end point was the improvement in tremor severity and activities of daily living (ADL), assessed by comparing preoperative and postoperative Fahn-Tolosa-Marin (FTM)¹⁶ scores. Treatment failure was defined as a relative improvement of <35%, whether a relative improvement >66% was classified as a “remarkable” outcome.

The secondary outcome measures were as follows: Tremor improvement was objectively assessed by comparing preoperative and postoperative accelerometry recordings, reported as tremor amplitude (mg²·s·Hz), see Supplementary Material (http://links.lww.com/NEU/F126) for details of the measurement. The impact on quality of life was evaluated based on the relative improvement in modified Parkinson's Disease Questionnaire (mPDQ-39)¹⁷ scores. All side effects and complications were recorded. A specific analysis of ataxia was conducted using the Scale for Assessment and Rating of Ataxia (SARA).¹⁸ In addition, in patients from center 1, balance was objectively assessed using static posturography (see Supplementary Material [http://links.lww.com/NEU/F126] for details of the measurement).

Target and Active Contact Locations

To better understand the spatial relationships between the predicted target (OptimDBS), the active contact and the thalamus, we aimed to localize the OptimDBS target, the active contact (its center), and the corresponding volume of tissue activated (VTA, see Supplementary Material [http://links.lww.com/NEU/F126] for calculation) within the anterior commissure–posterior commissure stereotactic space and in relation with thalamic anatomy derived from the Histogram-based Polynomial Synthesis-Thalamus Optimized Multi-Atlas Segmentation (HIPS-THOMAS¹⁹) procedure, an enhanced version of the THOMAS thalamic segmentation method²⁰ which uses 3D T1-weighted MRI data as input. The HIPS-THOMAS segmentation was merged with stereotactic MRI and postoperative computed tomography scans using the Lead-DBS software²¹ (www.lead-dbs.org). Then, we computed Euclidean distances (in mm; see Supplementary Material for the Mathematical Method, http://links.lww.com/NEU/F126) between the following structures: the OptimDBS target, the active contact (center and VTA boundary), the boundary, and the most ventral point of the ventral-intermediate nucleus of the thalamus (VIM) derived from the HIPS-THOMAS atlas. In addition, to assess the stereotactic accuracy of the surgical procedure, we calculated the distance between the OptimDBS target and the axis of the implanted DBS lead.

Statistical Analysis

The sample size of 22 patients was determined based on the hypothesis of a mean reduction of 66% (±15% SD) in FTM scores between preoperative and 3-month postoperative assessments. Assuming a bilateral first-species α risk of 5%, this sample size provided a statistical power exceeding 90% to demonstrate, using the Student t-test, that the reduction was greater than 55%—a threshold consistent with the mean improvement reported in the literature for DBS.²² Qualitative variables are reported as counts and percentages. Quantitative variables are expressed as median [first and third quartiles], mean ± SD, and mean (95% CI). Statistical analyses were conducted using SAS software v9.4 (SAS Institute) and R software (R Development Core Team).

Standard Protocol Approvals, Registrations, and Patient Consents

The study protocol was approved by national ethics committees. The study was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all patients.

Data availability

The study protocol and statistical analysis plan are available on request.

RESULTS

Between June 2019 and May 2023, 22 patients were prospectively enrolled in this trial. The inclusion period was extended by 2 years because of the suspension of functional neurosurgery activities during the COVID-19 pandemic. Follow-up was completed for all patients as per the protocol, with no losses to follow-up, missing data, or protocol deviations (Figure 1).

FIGURE 1.

Consolidated standards of reporting trials 2010 flow diagram.

General Characteristics of the Patients and Stimulation Parameters

The male-to-female sex ratio was 0.69. The median age at inclusion was 69 years [58; 71], with a median disease duration of 19 years [10; 31]. All enrolled patients underwent bilateral surgery, resulting in the implantation of 44 electrodes. The median surgical duration was 179 minutes [98; 210]. Stimulation was monopolar for all leads. The most frequently used contacts were the distal ones (#0: n = 15, 34%; #1: n = 16, 36%). The more proximal contacts (#2 and #3) were used in 5 (11%) and 8 cases (18%), respectively. Stimulation parameters (mean ± SD) were as follows: voltage, 1.9 ± 0.5 V; current, 1.56 ± 0.62 mA; pulse width, 60 ± 0 µs; and frequency, 130 ± 0 Hz, except for 1 active contact set at 170 Hz.

Efficacy on Tremor and Quality-of-Life Improvement

Table 1 presents the clinical assessments at baseline and 3 months postoperatively. The mean relative improvement in FTM scores between the preoperative and 3-month postoperative assessments was 61.3% ± 17.1% (95% CI: [53.7%; 68.9%]). One patient (4.5%) was classified as “treatment failure,” whereas 8 patients (36.4%) demonstrated a remarkable response to surgery. Between baseline and 3 months postoperatively, axial tremor (FTM items 1-4, 7, and 15) decreased by 82% ± 20%. Postural tremor on the right side was reduced by 75% ± 33% according to the FTM scale (items 5 and 8) and by 81% ± 56% based on accelerometry. On the left side, postural tremor decreased by 61% ± 36% according to the FTM scale (items 6 and 9) and by 84% ± 35% based on accelerometry. Action tremor (items 6, 9, and 11-14) decreased by 53% ± 26% on the right side and by 39% ± 45% on the left side. ADL (items 16 to 21 of the FTM) improved by a mean of 78% ± 18%. Quality of life assessed by the mPDQ-39 improved by a median of 55% (Q1-Q3: 24.3%-77.8%). These results are summarized in Figures 2 and 3. Medication use, particularly propranolol and primidone, was reduced.

TABLE 1.

Clinical Assessments Before and After Intervention (n = 22)

Variable	Baseline	3 mo postoperatively
Essential tremor
Fahn-Tolosa-Marin score (maximum: 144)	52 [44; 60]	21 [13; 26]
Patients receiving medical treatment	13 (59%)	10 (45%)
Propranolol	12	8
Primidone	3	1
Lamotrigine	1	1
Topiramate	1	1
Gabapentin	2	0
Clonazepam	1	1
Ataxia
SARA score (maximum: 40)	4 [3; 6]	3 [2; 6]
Quality of life
mPDQ39 score (maximum: 156)	38 [32; 55]	23 [6; 34]

mPDQ, modified Parkinson's Disease Questionnaire; SARA, Scale for the Assessment and Rating of Ataxia.

Values are median [Q1; Q3] or n (%).

FIGURE 2.

Preoperative and postoperative mean Fahn-Tolosa-Marin scores for postural and action extremity tremors (items 5, 6, and 8-14), axial tremor (items 1-4, 7, and 15), and ADL (items 16-21). ADL, activities of daily living.

FIGURE 3.

Amplitude (total accelerometric power, log [mg²·s·Hz]) of postural tremor on the left and right sides before and after surgery, recorded with arms abducted, elbows flexed, and the fingers of both hands pointing toward each other. Error bars: 95% CI.

Side Effects and Procedural Complications

Eighteen adverse events (AEs) were recorded during the study (Table 2), of which 10 events in 9 patients were attributable to the procedure tested in the study. Except for 1 grade 3 event—transient aphasia due to a transient ischemic stroke—all AEs were classified as grade 1 or 2.

TABLE 2.

Adverse Events

Adverse events
Number of AEs	18
Patients with at least 1 AE	12	(54.5%)
Grade
1	13 events in 9 patients	72.2%/40.9%
2	4 events in 3 patients	22.2%/13.6%
3	1 event in 1 patient	5.5%/4.5%
4	0	0
5	0	0
Imputabilitya	10 events in 9 patients	55.6%/40.9%
Type of AE
Self-reported ataxiab	5	22.7%
Wound healing disordersa	3	13.6%
Sinusitis	2	9%
Transient confusiona	1	4.5%
Transient aphasiaa	1	4.5%
Nausea	1	4.5%
Transient diplopia	1	4.5%
Fatigue	1	4.5%
Headache	1	4.5%
Carpal tunnel syndrome	1	4.5%
Renal failure	1	4.5%

AE, Adverse event.

^aProcedure-related AE.

^bStimulation-induced AE.

The SARA scores (mean ± SD) at baseline and 3 months postoperatively were 5 ± 3.3 and 3.9 ± 3, respectively. Figure 4 provides a detailed breakdown of the SARA section scores. Overall, we observed stability, with a tendency toward improvement across all items. Scores for items 1-3 (gait, stance, and sitting) improved in 9 patients (41%), remained stable in 8 patients (36%), and worsened in 5 patients (23%).

FIGURE 4.

SARA assessment of unsteadiness (items 1-3), dysarthria (item 4), dysmetria (items 5, 6, and 8), and adiadokokinesis (item 7). SARA, Scale for the Assessment and Rating of Ataxia.

Posturography analyses of the 13 patients from center #1 showed that, between baseline and 3 months postoperatively, center of pressure (CoP) displacement length decreased in 9 patients (69%) and increased in 4 patients (31%) under the eyes-open condition. In the eyes-closed condition, CoP displacement length decreased in 7 patients (54%) and increased in 6 patients (46%). CoP displacement area decreased in 5 patients (38%) and increased in 8 patients (62%) under the eyes-open condition, whereas in the eyes-closed condition, it decreased in 3 patients (23%) and increased in 10 patients (77%).

Location of the Target and DBS Lead

The stereotactic accuracy of the procedures, given by the shortest distance (mean ± SD) between the OptimDBS target and the DBS lead axis, was 1.47 ± 0.61 mm. The distance (mean ± SD) between the OptimDBS target and the active contact was 2.61 ± 1.18 mm. The VTA encompassed 25 OptimDBS targets (57%). The locations of the target and the active contact (including the VTA) for each patient are shown in Figure 5 and the Supplementary Material (Figure S2, http://links.lww.com/NEU/F126). Table 3 presents the stereotactic coordinates of the OptimDBS target and active contact, along with the shortest distances between the target, active contact, VTA, and VIM. The VTA reached the VIM and/or posterior subthalamic area (PSA) in all 44 DBS leads. The VTA reached only the VIM in 2 leads (4%), only the PSA in 10 leads (23%), and both the VIM and PSA in 32 leads (73%).

FIGURE 5.

Reconstructions of deep brain stimulation leads (n = 44) in Montreal Neurological Institute space (ICBM-15228) showing the ventral-intermediate nucleus of the thalamus (ventrolateral posterior nucleus in the Thalamus Optimized Multi-Atlas Segmentation Atlas, brown structure) and thalamus (red transparent structure).

TABLE 3.

Distances (in mm, Mean ± SD) Between the VIM Segmented by the HIPS-THOMAS Algorithm,^13,25 OptimDBS Target, and DBS Lead (Active Contact and VTA²⁴) (n = 44)

Origin point for distance calculation	Coordinates (PC at origin, mm)/VTA radius (mean ± SD, mm)	Distance to VIM boarder	Inside VIM	Distance to VIM ventral point	In PSAa
OptimDBS target	X = 13.63 ± 0.9 Y = 7.93 ± 1.16 Z = 1.49 ± 1.39	0.39 ± 0.84	31 (70%)	2.32 ± 1.13	10 (23%)
Active contact	X = 13.63 ± 1.75 Y = 7.7 ± 1.74 Z = 0.03 ± 1.81	1.42 ± 1.56	9 (20%)	2.8 ± 1.22	26 (59%)
VTA	2.66 ± 0.57	0.31 ± 0.80	34 (77%)	0.68 ± 1.01	42 (95%)

DBS, deep brain stimulation; HIPS-THOMAS, Histogram-based Polynomial Synthesis-Thalamus Optimized Multi-Atlas Segmentation; PC, posterior commissure; VIM, ventral-intermediate nucleus of the thalamus; VTA, volume of tissue activated.

^aZ coordinate of structure < Z coordinate of VIM ventral point.

DISCUSSION

This phase 2 study showed that asleep thalamic DBS, guided by the OptimDBS algorithm, is effective in treating ET, achieving a 61.3% reduction in FTM scores at 3 months. One patient improved by 34%, just below the treatment failure threshold, with a low baseline score at M0. This raises the methodological question of whether relative change is the most appropriate measure of benefit, as absolute change or final score thresholds may sometimes be more meaningful. The mean reduction in intention tremor amplitude, as measured by accelerometry, was over 80%. Asleep surgery reduces procedural complexity and operative time. In addition, direct implantation of the definitive lead, without previous intracerebral trajectory for microelectrode recordings, likely reduces intraoperative cerebrospinal fluid leak and pneumocephalus. This should reduce the risk of brain shift and make DBS lead implantation at the target more precise. Finally, asleep surgery could enhance accessibility for patients who cannot tolerate awake surgery.

This level of improvement aligns with both awake and asleep DBS outcomes reported in both prospective and retrospective studies by experienced DBS teams.^{4,5,7-9,23-43} Gravbrot et al⁸ and Wakim et al⁷ observed tremor suppression rates of 55%-65% with imaging-guided asleep DBS, whereas Paschen et al³⁷ reported long-term tremor control of approximately 60% with awake DBS. In a systematic review including 1093 patients treated between 1998 and 2018, Dallapiazza et al²² reported tremor control rates at 12 months ranging from 66% to 78% for bilateral stimulation and approximately 75% at long-term follow-up. Ablative approaches such as magnetic resonance–guided focused ultrasound and Gamma Knife radiosurgery thalamotomy yield comparable short-term outcomes, with studies reporting tremor reductions of 56%-60%.^44,45 However, comparisons with the literature is limited by the heterogeneity of the designs, the judgment criteria (overall assessment including ADL, as in our study, or simple assessment of tremor) and the follow-up times.

In addition, we showed a 78% improvement in the ADL and a 55% improvement of the quality of life. These benefits were particularly pronounced in precise motor tasks such as eating, writing, and pouring liquids—activities essential for maintaining daily independence. Comparable improvements have been reported in other DBS studies. For example, Persson et al⁴¹ observed substantial reductions in ADL scores, with improvements exceeding 70% at 1 year.⁴¹ In the systematic review by Dallapiazza et al,²² improvements in quality of life ranged from 57.9% to 82%.

No thalamic stimulation–induced ataxia was observed in our short-term study. However, individual assessments using a multimodal approach—including self-reports, items 1-3 of the SARA, and posturography (when available)—yielded conflicting results, with gait disturbances detected in 5-10 patients. Furthermore, stimulation-induced ataxia can develop progressively over the long term.⁴⁶ Therefore, we cannot conclude that our approach reduces the risk of ataxia compared with other methods. Previous reports have documented ataxia incidence rates of 15%-25% after DBS,^7,47 with ataxia rates below 10% in awake surgery.^37,48 By contrast, ablative techniques, such as magnetic resonance–guided focused ultrasound, are associated with higher ataxia rates ranging from 25% to 35%, as reported by Mohammed et al.⁴⁹

Although the mean Euclidean distance between the predicted target and the active contact was 2.6 mm, this metric must be interpreted within the context of known sources of spatial variance, including the electrode radius (0.6 mm), image coregistration errors (∼1.2 mm as previously quantified⁵⁰), and the stereotactic accuracy (∼1.5 mm in the study). The average radius VTA was 2.66 mm, which implies that although clinicians can fine-tune stimulation postoperatively, the lead must still be implanted within the correct anatomic zone for this flexibility to be meaningful. In other words, the capacity to “drive a VTA in the region” presupposes that the lead trajectory already intersects that effective region. This is the contribution of OptimDBS: It predicts accurately a patient-specific target that allows the electrode to be positioned within a therapeutically relevant volume encompassing the VIM and/or PSA.

The strengths of this study include its prospective design with standardized, multidimensional assessments of both efficacy and safety and the use of a reproducible, automated targeting approach based solely on conventional 3D T1-WI. In addition, the method used to analyze the anatomic relationship between the target, implanted electrodes and the VIM and PSA was based on one of the most advanced atlases available for intrathalamic nucleus segmentation.²⁰

Limitations

Despite these strengths, several limitations warrant discussion. The relatively small sample size of 22 patients limits the generalizability of our findings, particularly given the single-arm study design. This highlights the need for larger, multicenter, controlled studies to validate the results. Second, the 3-month follow-up period precludes an assessment of long-term outcomes, such as ataxia risk and the durability of functional improvements. Another limitation is the lack of connectome/tractography analyses, which could have refined electrode anatomic localization and elucidated associations with therapeutic efficacy and side effects, but such pathophysiological insights were beyond this study's scope.

CONCLUSION

Asleep DBS guided by the OptimDBS algorithm shows promising clinical outcomes comparable with those of awake procedures and ablative techniques. Larger controlled studies with longer follow-up are needed to confirm the results. This approach may also hold potential for application in lesioning therapies.

Supplementary Material

neu-99-324-s001.docx ^{(5.7MB, docx)}

Funding

This interventional study received funding from Bordeaux University Hospital's 2018 Internal Call for Tenders. None of the commercial companies mentioned in this study (Medtronic™ and RebrAIn™) contributed to the funding or any aspect of the trial, including its design, data collection, or interpretation of the results.

Disclosures

The OptimDBS algorithm was developed through academic research conducted at Bordeaux University Hospital, the University of Bordeaux, and INRIA research center. This research was subsequently transferred to RebrAIn, a startup spinoff of these institutions founded in 2021. Nejib Zemzemi and Emmanuel Cuny are co-founders and shareholders of this startup. Thomas Tourdias is a member of the Thalamic Nuclei Neuro-imaging Group consortium (https://thalamicsegmentation.github.io/) and acknowledges valuable scientific discussions regarding the THOMAS atlas. Thomas Tourdias is supported by the University of Bordeaux's IdEx Investments for the Future program, RRI IMPACT, and the IHU Precision & Global Vascular Brain Health Institute (VBHI), funded by the France 2030 initiative (ANR-23-IAHU-0001). The other authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

SUPPLEMENTAL DIGITAL CONTENT

Section I. Description of the OptimDBS algorithm.

Table S1. Landmark descriptions and locations.

Figure S1. The 18 anatomic landmarks on a T1-weighted MRI. A, Axial plane at the level of (AC-PC), B, sagittal plane at the level of landmark “A” (lateral border of the third ventricle), C, axial plane at the coordinate z = ZC – 5 mm, D, axial plane at the coordinate z = ZC, E, axial plane at the coordinate z = ZC – 10 mm, F, coronal plane at the level of landmark “PM1.”

Section II. Methods for the OPTIM-VIM study.

Section III. Supplemental results of lead location on the HIPS-THOMAS Atlas.

Figure S2. Reconstructed MRI (coronal plane) of the DBS lead, VTA, VIM segmented by the HIPS-THOMAS algorithm, and the OptimDBS target (white cross).

COMMENTS

We congratulate the authors on their well-organized work reporting the results of a prospective, unblinded, uncontrolled, phase II clinical trial. The work evaluates the safety and effectiveness of a novel technology for automatic targeting stereotactic interventions for tremor management in a cohort of 22 adult patients (44 leads) with medically refractory essential tremor. Overall, the outcomes seem to be consistent with other cohorts of patients undergoing DBS for tremor and the most common complication was gait disturbance. Because of the design, and lack of blinding and controls, it is not possible at this stage to distinguish how much of the benefits and adverse effects are related to the automated algorithm or the manual intervention of the surgeons. While the work represents an important milestone in the long-term process of testing a new technology, some limitations can be considered. The sample size is limited and accrued over a relatively long period of time, which could bias the results. The follow-up data at three months is short, and we hope the long-term results will be published in subsequent reports by this group.

Andre Machado

Cleveland, Ohio, USA