Digital Architecture for MUAPs Propagation Speed Estimator triggered by Foot Plant Switch

In this paper, we propose the design and the FPGA implementation of a wireless and wearable estimator of the motor unit action potentials (MUAPs) propagation speed by using footswitches as trigger. The system evaluates the propagation speed during dynamic muscles contractions like during the walking. Starting from our previous design that exploits 4 wireless EMGs placed on the lower limbs and a single XNOR gate cross-correlation method to extract the MUAPs propagation speed (MPS), we implement an FPGA-based plugin able to improve the MPS readings stability and repeatability. This measurement stabilizer exploits 4 Force-Sensitive Resistors (FSRs) per foot plantar and a cascade of comparators to define a pattern of foot plant involvement during the gait. It is used to activate the MPS estimation unit and the external transmission of the data. The proposed FPGA-implemented plug-in adds only 125 ns to the previously recorded time delay of 370 ms, to complete the processing chain preserving the real-time operation. At the same time, it increases the resource utilization of only 35/32070 ALMs, 36 ALUTs, 41/64140 registers. The invivo tests on six subjects showed an improvement in the MPS estimates standard deviation of the 22.5% with a triggered MPS estimator with respect to a not-triggered one (i.e., from ±0.36m/s to ±0.28m/s). The repeatability test also demonstrates that the outliers' presence is reduced, on average, of the 1.5% if the system is triggered (+14.5% in the best-case study).