ISSUE 003 · Oct 5, 2026 · 8 min read

What if deep brain stimulation could follow the rhythm of gait?

Detect emerging gait freezing from wearable and neural signals, then adjust stimulation by gait phase inside clinician-defined safety bounds.

The issue in one sentence

Detect emerging gait freezing from wearable and neural signals, then adjust stimulation by gait phase inside clinician-defined safety bounds.

Diagram 01

A closed loop paced by gait state

01Step · posture→

02Neural sensing→

03Gait-phase estimate→

04Freeze risk→

05Safety gate→

06Stimulation→

07Falls · function

ANM proposed model. It connects approved neural-signal adaptive DBS with a 2026 gait-synchronization feasibility study while keeping the gait algorithm explicitly investigational.

01Clinical problem · Evidence

Freezing and falls can persist even when other motor symptoms improve.

Gait dysfunction reduces independence and quality of life. Conventional DBS delivers continuous stimulation, while gait changes phase by phase and its long-term gait benefit may diminish.

02Technology stack · Proposal

Put neural activity and each step on the same timeline.

Sensing-enabled DBS records local field potentials while ankle and waist IMUs estimate gait phase and possible freezing. AI fuses both signals to time amplitude changes, constrained by on-device limits and immediate reversion to baseline.

03Research question · Hypothesis

Does personalized gait-phase stimulation reduce real-world freezing versus continuous DBS?

Test whether it reduces real-world time spent freezing and near-falls without worsening tremor, rigidity, speech, or overall motor control.

04First study · Proposal

36 participants, double-blind crossover, six weeks per condition.

Randomize people with stable medication and implanted bilateral sensing DBS plus freezing of gait to optimized continuous DBS and gait-synchronized aDBS in counterbalanced order. The primary outcome is wearable-derived real-world freezing proportion; key safety outcomes are falls and stimulation-related adverse events. Also measure MDS-UPDRS III, speech, battery use, and preference.

05Defensible data · Venture assumption

Link each stimulated step to a clinical outcome.

Align neural signals, gait phase, stimulation changes, freezing, falls, and medication timing for each patient. The defensible asset is intervention-to-outcome data showing which bounded settings worked safely for whom—not a gait classifier alone.

06Venture thesis · Assumption

The first product is a programming episode, not an autonomous algorithm.

Initial customers are high-volume movement-disorder centers and device-company clinical or postmarket teams. Bundle a seven-day wearable assessment, safety-envelope design, in-person programming, and four-week outcome report per episode; later test implanted-patient-year subscriptions and trial analytics.

Evidence map

The gait gap between authorization and clinical value

Feb 2025

FDA PMA supplement approval of an optional adaptive DBS feature

68 people

Enrollment in the pivotal long-term at-home ADAPT-PD trial

−15%

Mean change in delivered electrical energy with single-threshold aDBS versus continuous DBS

5 people

Size of the gait-synchronized aDBS study; three completed the at-home crossover

Evidence

What is established—and what is not

Established: FDA approved an optional adaptive feature for an existing DBS system in February 2025. In the 68-person ADAPT-PD study, both aDBS modes met the prespecified performance goal, and single-threshold aDBS delivered 15% less electrical energy on average than continuous DBS. Participants were already stable on continuous DBS and medication, and the study was manufacturer-supported. A 2026 gait-synchronized study identified personalized gait-phase biomarkers in five people, but only three completed the at-home crossover. Reduced falls and broad generalizability therefore remain hypotheses.

FDA PMA supplement ADAPT-PD pivotal trial Gait-synchronized feasibility trial

Failure conditions and safety, ethics, and regulatory risks

Clinical failure condition

Do not claim superiority unless real-world freezing falls by at least 20% without worsening falls, speech, or overall motor control.

Immediate fallback

On signal loss, sensor detachment, or out-of-distribution gait, stop adaptation and revert immediately to the clinician-validated continuous setting.

Stimulation harm

Gait gains can coexist with dyskinesia, impaired speech, or worse balance. Predefine patient-specific amplitude bounds, ramp rates, and daily exposure.

Representativeness

Selecting only patients with clean signals and easy wearable use limits generalizability. Report cognitive impairment, mobility-aid use, and diverse gait phenotypes separately.

Regulatory boundary

The 2025 approval covers a specific neural-signal adaptive feature. A gait-phase algorithm and new stimulation policy are not automatically covered and may require an IDE and PMA supplement.

Business failure condition

Do not scale if total programming, visit, and training costs exceed standard DBS without patient-level improvement in freezing or falls.

Bibliography

References

Established facts are separated from research and venture hypotheses that still require validation.

01
U.S. Food and Drug Administration. Activa®, Percept™ and SenSight™ Deep Brain Stimulation Therapy System. PMA P960009/S478.
Primary regulatory record for the approval date and indication of the optional adaptive DBS programming feature.P960009/S478 · Feb 20, 2025
02
Bronte-Stewart HM, Beudel M, Ostrem JL, et al. Long-Term Personalized Adaptive Deep Brain Stimulation in Parkinson Disease: A Nonrandomized Clinical Trial. JAMA Neurol. 2025;82(11):1171–1180.
A 68-person international pivotal trial supporting performance and energy findings, but not a direct test of gait benefit versus continuous DBS.PMID 40982287 · NCT04547712
03
Oehrn CR, Cernera S, Hammer LH, et al. Chronic adaptive deep brain stimulation versus conventional stimulation in Parkinson’s disease: a blinded randomized feasibility trial. Nat Med. 2024;30:3345–3356.
A four-person personalized aDBS feasibility study demonstrating symptom-specific biomarker selection.PMID 39160351
04
Louie KH, Balakid JP, Bath JE, et al. Adaptive deep brain stimulation for dynamic gait control in Parkinson’s disease: a randomized feasibility trial. Nat Med. 2026;32:2803–2814.
Identified gait-phase biomarkers in five participants; three completed the at-home crossover. Fall reduction is exploratory and requires confirmation.PMID 42297980 · NCT04675398
05
ClinicalTrials.gov. Adaptive Deep Brain Stimulation for Freezing of Gait in Parkinson’s Disease.
Registry evidence that adaptive DBS specifically targeting freezing of gait is under investigation.NCT06819020

The next issue publishes every day at 9 AM KST.All issues