Bronte-Stewart Lab

Brain 144(2):473–486 (online December 2020) · 2021

Modulation of beta bursts in subthalamic sensorimotor circuits predicts improvement in bradykinesia

Kehnemouyi YM, Wilkins KB, Anidi CM, Anderson RW, Afzal MF, et al., Bronte-Stewart HM

Why this matters: Relates stimulation-associated changes in beta bursts to improved bradykinesia. The primer uses abstract-level evidence; an association does not establish a sole cause.

DOI: 10.1093/brain/awaa394

Brain Commun 7(4):fcaf266 · 2025

Beta burst-driven adaptive deep brain stimulation for gait impairment and freezing of gait in Parkinson's disease

Wilkins KB, Petrucci MN, Lambert EF, Melbourne JA, Gala AS, Akella P, Parisi L, Cui C, Kehnemouyi YM, Hoffman SL, Aditham S, Diep C, Dorris HJ, Parker JE, Herron JA, Bronte-Stewart HM

Why this matters: Seven-participant investigational beta-burst-driven adaptive DBS study. Gait/freezing improved relative to stimulation OFF; group outcomes were comparable to continuous and random-adaptive stimulation, without universal superiority.

DOI: 10.1093/braincomms/fcaf266

npj Digit Med 8:7 · 2025

N2GNet tracks gait performance from subthalamic neural signals in Parkinson's disease

Choi JW, Cui C, Wilkins KB, Bronte-Stewart HM

Why this matters: N2GNet estimated gait from implanted STN LFPs in 18 participants, with later-visit testing and stimulation OFF. Offline prediction is not proof of mechanism or a validated therapy controller.

DOI: 10.1038/s41746-024-01364-6

Brain Commun 8(4):fcag245 · 2026

Reduced pre-movement subthalamic beta desynchronization marks motor deficit in Parkinson's disease

Seo G, Wilkins KB, Bronte-Stewart HM

Why this matters: In 16 participants, reduced pre-movement beta desynchronization was associated with worse motor performance across three tasks. This is distinct from beta power or burst duration.

DOI: 10.1093/braincomms/fcag245

Ann Neurol 93(5):1029–1039 · 2023

Bradykinesia and its progression are related to inter-hemispheric beta coherence

Wilkins KB, Kehnemouyi YM, Petrucci MN, Anderson RW, Parker JE, Trager MH, et al., Bronte-Stewart HM

Why this matters: Links bradykinesia and its progression with interhemispheric beta coherence over longitudinal observation. Coherence between signals differs from the power of a single signal; association is not causation.

DOI: 10.1002/ana.26605

Brain Stimul 19(1):103028 (letter / case report) · 2026

At home monitoring of chronic adaptive deep brain stimulation for Parkinson's disease

Cui C, Choi JW, Karjagi S, Wilkins KB, Negi A, Bronte-Stewart HM

Why this matters: One-person at-home case report compares four-week adaptive and continuous DBS periods with neural sensing and daily digitography. Feasibility and individual patterns cannot establish population-level superiority.

DOI: 10.1016/j.brs.2026.103028

J Parkinsons Dis 12(6):1979–1990 · 2022

Quantitative digitography measures motor symptoms and disease progression in Parkinson's disease

Wilkins KB, Petrucci MN, Kehnemouyi Y, Velisar A, Han K, Orthlieb G, Trager MH, O'Day JJ, Aditham S, Bronte-Stewart H

Why this matters: Quantitative digitography measured tapping timing, amplitude and rhythm in 96 people with Parkinson’s and 42 controls. Instrumented measurements complement clinical ratings without replacing them.

DOI: 10.3233/JPD-223264

J NeuroEng Rehabil 19:20 · 2022

Assessing inertial measurement unit locations for freezing of gait detection and patient preference

O'Day J, Lee M, Seagers K, Hoffman S, Jih-Schiff A, Kidziński Ł, Delp S, Bronte-Stewart H

Why this matters: Evaluates wearable IMU locations for freezing detection and patient preference. Sensor placement and validation matter; detection alone is not evidence that treatment works.

DOI: 10.1186/s12984-022-00992-x

Introductory

Detect candidate freezing events from wearable IMU data

Computational / machine learning and signal processing

Compare a threshold baseline and a classifier on public or synthetic gait signals. Separate people between training and testing and report errors around turning.

Background

Gait, IMUs, Model evaluation

Data

Synthetic IMU traces, or a publicly licensed gait dataset with documented consent and de-identification. No private clinical recordings.

Output

Reproducible notebook, plots and a short claim-evidence report

Independent educational idea, not offered or supervised by the lab. Public or synthetic data only; no human-subject, animal or clinical intervention. Simulated controllers must never be used on a person or stimulation device.

Introductory

Detect beta bursts in simulated neural recordings

Simulation / neural signal analysis

Generate a declared beta-band signal with known transient bursts. Test how filtering and threshold choices change estimated burst duration.

Background

Beta oscillations, Python

Data

Synthetic 13–30 Hz signals with an explicit burst-duration ground truth.

Output

Reproducible notebook, plots and a short claim-evidence report

Independent educational idea, not offered or supervised by the lab. Public or synthetic data only; no human-subject, animal or clinical intervention. Simulated controllers must never be used on a person or stimulation device.

Introductory

Compare power, burst duration and coherence

Computational neuroscience / literature-data synthesis

Simulate pairs of signals where average power stays fixed but bursts or intersignal coherence change. Compare distinct measurements with an evidence table from cited papers.

Background

Beta bursts, Statistics

Data

Synthetic signal pairs and publicly reported summary values from the cited beta studies.

Output

Reproducible notebook, plots and a short claim-evidence report

Independent educational idea, not offered or supervised by the lab. Public or synthetic data only; no human-subject, animal or clinical intervention. Simulated controllers must never be used on a person or stimulation device.

Introductory

Simulate open-loop and feedback stimulation

Simulation / control systems modeling

Compare a fixed controller with a bounded feedback policy in a toy noisy motor-state model. Report delay sensitivity and failure modes without making clinical predictions.

Background

Adaptive DBS, Python

Data

A synthetic plant model with explicitly hypothetical parameters; no device connection.

Output

Reproducible notebook, plots and a short claim-evidence report

Independent educational idea, not offered or supervised by the lab. Public or synthetic data only; no human-subject, animal or clinical intervention. Simulated controllers must never be used on a person or stimulation device.

Introductory

Quantify finger-tapping variability

Computational / kinematics and statistics

Analyze synthetic alternating-key timing and amplitude traces. Compare mean speed with rhythm variability and show how measurement noise changes conclusions.

Background

Quantitative digitography, Statistics

Data

Synthetic finger-tapping traces; reported QDG summary values for context only.

Output

Reproducible notebook, plots and a short claim-evidence report

Independent educational idea, not offered or supervised by the lab. Public or synthetic data only; no human-subject, animal or clinical intervention. Simulated controllers must never be used on a person or stimulation device.