Does beta ERS (event-related synchronization) cause motor recovery, or does it just track it? The post-movement rebound in beta power has been linked to motor learning, stroke rehabilitation, and Parkinson’s disease for years, but causality has been impossible to prove.
A new Oxford study gives the first causal answer. Using a closed-loop tACS system (transcranial alternating current stimulation) triggered by behavior, the team shows that boosting beta ERS directly improves motor retention.
How the closed-loop system was built
24 healthy adults completed three sessions at least a week apart. The study was double-blind, sham-controlled, and within-subject. Neither the participant nor the experimenter knew which condition was active.
EEG was recorded using mbt’s Smarting PRO X 64-channel wireless EEG device. The wireless form factor was essential here. Participants were using a robotic arm, and a wired EEG system would have caused cable artefacts and limited movement.
EEG signals were streamed with kinematic data from a VIVE Tracker 3.0 mounted on the robotic handle. That tracker provided the movement offset signal used to trigger tACS in real time.
Stimulation was delivered via a Starstim 8 system (Neuroelectrics), with an electrode over left motor cortex (C3) and a return at Pz.
Experimental setup and stimulation protocol
The task used a Kinarm END-Point robotic arm. Participants made outward and inward reaches between two targets. During the adaptation phase, a lateral force field was applied at varying levels. Participants weren’t told about it and had to adapt without feedback.
Right after, the force was switched off and visual feedback of the cursor was removed mid-movement. Participants just had to reach and see where they ended up – this revealed whether the compensation they had learned was still there, or had faded.
Throughout the task, tACS was delivered to the scalp over left motor cortex after each reach. Three conditions were tested across sessions.
In the early condition, 20 Hz stimulation ran for 2 seconds immediately after the hand stopped moving. In the late condition, the same stimulation arrived 2.5 seconds after movement end. In sham, a 100 ms pulse was delivered at the same delay as late – long enough to produce the same skin sensations, but too brief to have a meaningful neural effect.
Trials between stimulation were left unstimulated, and only those were used for EEG analysis.
The schematic below shows the 3 conditions timed relative to movement. Stimulation was anchored to a 150 ms dwell period in the target (t₁), detected by the VIVE Tracker.
Experimental setup and stimulation protocol. Panel A shows the force field sequence across the adaptation phase. Panel B: Example movement trajectories from a single participant at different force field magnitudes (0, 3, 6, and 9). Panel C shows stimulation timing. Early stimulation (orange) fired at t₁ for 2 seconds. Late (purple) fired 2.5 seconds later. Sham (blue) used the same trigger as late but lasted only 100 ms. This design let each timed window be tested against a matched control.Beta ERS goes up and ERD stays flat
Early tACS enhanced beta ERS over the contralateral sensorimotor cortex in the early post-movement window. Late tACS enhanced it in the later window. Each condition boosted beta power in the window it was designed to target. Importantly, all analyses used non-stimulated trials only, so the power changes reflect real shifts in brain state and not tACS artefact.
The timing of stimulation determines where beta ERS is enhanced. Panel A shows difference maps between each active condition and sham, with significant clusters outlined. Panel B traces the beta power time course for each comparison, with horizontal bars marking windows where the difference reached significance. Panel C shows mean beta ERS extracted from each window as box plots with individual participant data – early stimulation boosted beta in the early window, late stimulation in the late window.The beta ERD (power loss during movement itself) wasn’t affected by any condition. This is an important control. The stimulation was not raising cortex excitability broadly, but was doing something specific to the post-movement state. That selectivity matters for the causal claim.
Does boosting beta ERS improve motor retention?
Motor adaptation relies on two parallel processes. A fast process adapts quickly but fades – it’s thought to be cerebellar. A slow process learns more gradually but holds – it reflects cortical storage. The team fitted a two-state model to the trial-by-trial data to separate these.
Both early and late tACS raised the slow retention parameter versus sham. No other model parameter changed. The stimulation didn’t affect how fast participants adapted, but it affected only how well the motor cortex held onto what it had learned.
Panel A shows how lateral deviation changed trial by trial during the task, with model predictions (black lines) overlaid on the measured data (colored lines) for each condition. Panel B shows the four estimated model parameters as box plots; only the slow retention factor was higher in both tACS conditions compared to sham. The other three parameters were unchanged.The model result was confirmed by a simpler direct measure. When the force was removed and participants reached without seeing the cursor, those in the early tACS condition still compensated – their hand moved in the direction they had learned to correct for. They held onto the learning better than those in the sham condition.
The team also looked at whether individual differences in beta ERS predicted how much each person retained. Participants with stronger post-movement beta ERS tended to show better retention. The more beta ERS increased, the more the learning stuck.
What this means beyond the lab
Most tACS studies run continuous stimulation for 20 to 30 minutes. This study used brief 2-second bursts applied after each movement, totaling around 8.7 minutes per session. That is a much lower dose, yet it worked. The reason seems to be timing. Delivering stimulation at the moment the brain is already generating a beta rebound appears to be far more effective than applying the same current continuously throughout a session.
The system was also built with portability in mind. mbt’s wireless EEG removes the need for a fixed recording setup, and the VIVE Tracker is cheap and does not require a lab. Combined with a compact stimulator, the whole system could in principle be used outside a clinical setting.
The authors explicitly frame this as a step toward home-based rehabilitation for people with stroke or Parkinson’s disease, where beta ERS is persistently reduced.
Read the full paper linked here.
Curious if this setup is right for your study?
Our team works with researchers across neuroscience, sports science and clinical applications - we'll help you figure out if mobile EEG fits your study design.
