Development of tools for post-Stroke data collection - validation of novel fabric EMG sensor, with Arduino-driven data collection, on non-affected participants

BACKGROUND

Analysis of EMG signals when walking can potentially aid prescription of assistive devices such as Ankle Foot Orthoses (AFOs). In the Post-Stroke community AFOs are commonly prescribed to improve safe ambulation and prevent trips and falls [1]. However, current EMG sensors do not allow for accurate in-brace analysis due to Orthosis design and interaction with the sensors. Additionally, real-world data collection is limited.

AIM

Validation of novel fabric EMG sensor [2] combined with an Arduino-driven sensor, Myoware EMG board and Force Sensitive Resistor (FSR) foot switch (FS) for use with an assistive AFO.

METHOD

Data were collected from a convenience sample of non-affected participants wearing the novel fabric EMG device plus an adjustable AFO (GEO AFO with Triple Action Ankle joint, supplied by Becker Orthopaedic [3]. Spatiotemporal, kinematic and kinetic data were also collected through Vicon using the standard Plug in Gait (PIG) marker placement and modelling. Participants completed a minimum of 8x 5M walks in different AFO constraint configurations. Two different configurations have been reported here: 1) Novel fabric EMG device plus AFO set to ‘minimum’ ankle joint; 2) Novel fabric EMG device plus AFO set to 900 fixed ankle joint.

RESULTS

Data were gathered from 6 volunteers. 4 male (2 female), average age 26(+/- 4 years). AFO and fabric EMG was worn on the left leg for all participants. Figure 1 shows Arduino data collection device, Myoware board and FSR with fabric EMG reference electrode. FSR foot switch (FS) and EMG data were available for participants 1, 2, 4, 5 and 6. FS data for participant 3 was erroneous due to fault in FSR. EMG data were analysed in post-collection to give output in Volts (V) for comparison to existing published data. Data for each participant showed distinct peaks (n=4 (+/- 2)) for each participant throughout the collected range of 5M walked. FS data indicated distinct periods of contact vs no-contact comparable with stance and swing phases of gait.

DISCUSSION AND CONCLUSION

Timing of muscle activity was appropriate with standard gait cycle events as expected in non-affected participants. Spatiotemporal data showed slight trend towards increased contralateral stride length with the AFO in fixed 900 state. Addition of a forefoot FSR was indicated to improve gait event prediction. Corresponding EMG patterning and FS data indicated potential for the device as an analysis tool for gait.

REFERENCES

1. Mulroy, S. J., Eberly, V. J., Gronely, J. K., Weiss, W., & Newsam, C. J. (2010). Effect of AFO design on walking after stroke: Impact of ankle plantar flexion contracture. Prosthetics and Orthotics International, 34(3), 277-292. doi:10.3109/03093646.2010.501512 2. Footfalls&Heartbeats. (2020). 3. Becker. (2022). GEO AFO. Retrieved from https://beckerorthopedic.com/Product/AnkleComponents/TripleAction/3B00-GEO ACKNOWLEDGEMENTS Thanks to Footfalls and Heartbeats and Becker Orthopaedic. Research funded by the EPSRC through the CDT in Prosthetics and Orthotics.