Desarrollo de una herramienta computacional que permita la clasificación entre personas que presenten o no una patología a partir de señales electromiográficas de la extremidad inferior usando máquinas de soporte de vectores.

Abstract

This thesis proposes a tool for studying the relationship of surface electromyography between healthy people and someone with some pathology in the lower limb. Support Vector Machine (SVM) is used to classify electromyographic signals because models are robust to overfitting. For the lower limb, analysis has been taken the EMG Dataset from UCI Machine Learning. This database contains signals from 11 subjects with knee abnormality and 11 normally, previously diagnosed by a professional. They undergo three movements to analyze the behavior associated with the lower limb, gait, leg extension from a sitting position, and flexion of the leg up. Analysis has stages as preprocessing, features extraction, training, and validation. Several conventional electromyography features are used in performance comparison with it combined with Enhanced features. The training process was accomplished with 60% of the data and the validation of the algorithm with 40% of them. Finally, from the confusion matrix, evaluated from the final models, metrics such as Accuracy, Sensitivity, Specificity and Precision were calculated to select the model with optimal performance in real life.