Gabriel Graffagnino, Benoît Sijobert, Karine Patte, David Gasq, Christine Azevedo-Coste
The proposed approach enables reliable real-time gait-event detection and electrical stimulation triggering with latencies compatible with future clinical application. These results highlight the potential of using gait analysis platforms to flexibly explore and optimize personalized multichannel stimulation strategies, while assessing their immediate orthotic effects prior to translation of the most effective strategies into wearable systems.
BACKGROUND: Functional electrical stimulation (FES) can improve gait in children with cerebral palsy (CP), but gait impairments vary widely between individuals and likely require stimulation of different muscle groups. Delivering real-time, multichannel, gait-triggered stimulation remains difficult to implement in clinical practice. This study explores the technical feasibility of using gait analysis systems commonly available in pediatric rehabilitation centers to drive multichannel FES. The temporal performance of a real-time, gait-triggered system based on kinematic event detection was evaluated.
METHODS: Retrospective gait data from 30 children with CP were replayed to emulate real-time acquisition. A custom algorithm detected gait events (initial contact and a knee flexion threshold) and triggered an eight-channel stimulator whose output was recorded via an acquisition card through resistors. Hardware delays were measured across each stage: triggering function, command transmission, and stimulation onset, as well as overall system latency.
RESULTS: A total of 1100 gait cycles were analyzed. The mean global latency was 50.33 ms (SD 7.55 ms). Most stimulations occurred within 70 ms (99.45%), and all within 100 ms. The main source of delay was command transmission to the stimulator, while algorithm processing contributed minimally. Performance was consistent across triggering conditions.
CONCLUSION: The proposed approach enables reliable real-time gait-event detection and electrical stimulation triggering with latencies compatible with future clinical application. These results highlight the potential of using gait analysis platforms to flexibly explore and optimize personalized multichannel stimulation strategies, while assessing their immediate orthotic effects prior to translation of the most effective strategies into wearable systems.