Análisis de la presión plantar en amputados transtibiales unilaterales mediante una metodología ETL: Un enfoque para la clasificación de estados angulares del encaje de la prótesis

Abstract

The problem that is sought to be addressed is the difficulty of analyzing data from pressure sensors. plantar in individuals with unilateral transtibial amputation, making determination of alignment difficult adequate angle of the prosthesis socket. In order to respond to the concern, databases obtained when the individual is in static standing position with different angular locations of the prosthesis socket, using a ETL methodology to clean, transform, load and analyze the collected data. Three classification models will be developed using machine learning techniques to determine the different anatomical states of the prosthesis. These models will be evaluated to select the one with the best performance using relevant metrics. The methodology to be developed will have an important impact in the field of biomechanics and engineering of prosthetics by providing a robust tool for sensor data analysis in segment prosthetics lower body.