Evaluation of Machine Learning algorithms applied to differentiate upper extremities from bioimpedance measurements

GERARDO AMES-LASTRA; KATAISHI, RODRIGO; ALBERTO CONCU

CDMX

Conferencia; International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE); 2023

UNAM-XCDMX

Bioimpedance measurements provide every day more information about the physiological functions and the anatomical state in applications on humans. In general, each of the applications on human beings differs in the configuration of the technique used and in the processing of the data. At the same time, Machine Learning techniques are increasingly used in various biomedical fields. This paper analyzes the application of different Machine Learning algorithms to differentiate dominant arms and non-dominant arms from bioimpedance spectroscopy measurements in bipolar configuration and in a non-invasive way. In this work, we use the same data as in a previous work by our research group but with a different preconditioning and a reduced bioimpedance spectrum data size. The metrics obtained are promising and the methodology is suitable to include in studies focused on both arms of the subjects, as it reduces the bias introduced by the anatophysiological development of one dominant extremity over the other.