Voluntary movements are managed by movement related potentials (MRPs) which are brain activity patterns detectable even 500ms before the movement itself. The cortico-muscular matching between brain (EEG) and muscles (EMG) activity allows the assessment of the intentionality of the performed movement. Basing on this knowledge, a real-time algorithm for falling risk prediction based on EMG/EEG coupled analysis is presented. The system architecture involves 8 EMG (limbs) and 8 EEG (motor-cortex) channels wirelessly collected by a FPGA (gateway) that contextually performs the real-time processing based on an event triggered time-frequency approach. The digital architecture is validated on the FPGA to determine resources utilization, related timing constraints and performance figures of a dedicated real-time ASIC implementation for wearable applications. The system resource utilization is 85.95% ALMs, 43283 ALUTs, 73.0% registers, 9.9% block memory of an Altera Cyclone V FPGA. The processing latency is lower than 1ms and the output are available in 56ms, respecting the time limit of 300ms. Outputs enables decision-taking for feedback delivering.
The Truth Machine of Involuntary Movement: FPGA Based Cortico-Muscular Analysis for Fall Prevention / Annese, V. F; DE VENUTO, Daniela. - (2016), pp. 553-558. (Intervento presentato al convegno 15th IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2015 tenutosi a Abu Dhabi, UAE nel December 7-10, 2015) [10.1109/ISSPIT.2015.7394398].
The Truth Machine of Involuntary Movement: FPGA Based Cortico-Muscular Analysis for Fall Prevention
DE VENUTO, Daniela
2016-01-01
Abstract
Voluntary movements are managed by movement related potentials (MRPs) which are brain activity patterns detectable even 500ms before the movement itself. The cortico-muscular matching between brain (EEG) and muscles (EMG) activity allows the assessment of the intentionality of the performed movement. Basing on this knowledge, a real-time algorithm for falling risk prediction based on EMG/EEG coupled analysis is presented. The system architecture involves 8 EMG (limbs) and 8 EEG (motor-cortex) channels wirelessly collected by a FPGA (gateway) that contextually performs the real-time processing based on an event triggered time-frequency approach. The digital architecture is validated on the FPGA to determine resources utilization, related timing constraints and performance figures of a dedicated real-time ASIC implementation for wearable applications. The system resource utilization is 85.95% ALMs, 43283 ALUTs, 73.0% registers, 9.9% block memory of an Altera Cyclone V FPGA. The processing latency is lower than 1ms and the output are available in 56ms, respecting the time limit of 300ms. Outputs enables decision-taking for feedback delivering.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
