Sistema de clasificación SVM de señales electromiográficas extraídas en un sistema embebido

Abstract

The present work presents the design of a wireless electromyographic biomedical signal acquisition system, which records the muscle signals in the EKG / EMG development card, the signals are transmitted through the ZigBee protocol in point-to-point or multipoint link, so it is scalable for more than one patient in parallel. The transmission of the data is received in the Raspberry Pi3 development card which truncates the received signal and is sent to the cloud for a classification process. The developed system is a precise proposal of low cost for the analysis of several patients, the proposed technique represents the stage of acquisition, analysis and truncation of data for a signal classification process based on support of vector machines (SVM) with the In order to predict the best type of therapy for a given patient. Experimental and simulation tests developed in hardware and classified in software through SVM show that the complete