Within the framework of collaborative work between the University of Cauca and the Mar\u00eda Cano University Foundation, a study was carried out that proposes a multiclass model to classify the severity of knee osteoarthritis (OAR) using bioimpedance measurements. An experimental setup was implemented that included three types of measurements: global impedance with adjacent pattern, global impedance with opposite pattern, and direct impedance measurement. These measurements were performed using an electronic device proposed by the authors, based on the AD5933 impedance converter from Analog Devices.<\/p>\n
Data from the 15 experimental setups (3 types of measurements \u00d7 5 exercises) were used to train a multi-class random forest. The results showed that the proposed approach achieved mean sensitivities and specificities of 100% for the four degrees of OAR severity in the extension, cyclic flexion-extension, and walking tasks. In conclusion, this study suggests that the proposed method can serve as a screening tool to determine which individuals should undergo other imaging techniques, such as x-rays or magnetic resonance imaging, contributing to solving the challenge that represents the evaluation, diagnosis and follow-up of the evolution of the OAR.<\/p>\n
The main findings of the study have been published in the scientific article titled \u201cMachine learning-based bioimpedance assessment of knee osteoarthritis severity\u201d in the international and multidisciplinary journal Biomedical Physics & Engineering Express (DOI:10.1088\/2057-1976\/ad43ef), categorized in the SJR in Q3.<\/p>\n