Can foot movements or voice commands trigger electrical stimulation that helps stroke survivors grasp objects?

Details

Stroke is the leading cause of acquired motor deficiencies in adults. Restoring grasping abilities is challenging for people who have not recovered active hand opening after rehabilitation. A grasp neuroprosthesis (GNP) uses self-triggered functional electrical stimulation (FES) of finger extensor muscles to restore grasping in daily life, but remains poorly accessible to people after stroke. Researchers developed a GNP prototype with self-triggering controls adapted to this population and assessed its impact on abilities.

Across two clinical research protocols, 22 stroke participants used the GNP and its controls for 3 to 5 sessions over one week. Controls included electromyographic (EMG) activity of a predefined muscle, inertial measurement unit (IMU) motion detection, foot switches and voice commands. NeuroPrehens software interpreted commands from electromyographic, inertial, foot-switch or microphone sensors to trigger an external electrical stimulator using two bipolar channels with surface electrodes.

Participants tested 9 control modalities, subjectively rated their ease of use and reliability out of 10, and selected a preferred control. They then trained with the GNP to perform standardized, functional, one-handed grasping tasks while seated. Responsiveness and functional impact were assessed by retrospective analysis of task video recordings from two blinded-evaluation, multi-crossover N-of-1 randomized controlled trials.

Triggering with the non-paretic foot, using either EMG or IMU, received the highest ratings. Median ease-of-use scores out of 10 were 10 for EMG and 9 for IMU; median reliability scores were 9 for both. These controls were viable and appreciated by users, as were voice control and sideways head-tilt controls.

Combining the system's overall responsiveness with control-specific latencies, the median time from user intent to FES triggering was 333 ms for IMU controls (95% confidence interval: 211 to 561 ms), 217 ms for EMG controls (167 to 355 ms), and 467 ms for voice controls (147 to 728 ms).

The abstract reports significant functional improvement with GNP use in the two grasping tasks evaluated, with a median improvement of 3 points out of 5 (95% confidence interval: -1 to 5 points).

The researchers concluded that the prototype and its controls were well suited to people after stroke in terms of self-triggering, responsiveness and restoration of functional grasping abilities. A wearable version is being developed to improve grasping abilities at home.

Both studies are registered on clinicaltrials.gov: NCT03946488, registered May 10, 2019, and NCT04804384, registered March 18, 2021.