Machine learning methods for image processing in plantography method for physical rehabilitation

Plantar pressure distribution provides insights into foot function, gait mechanics, and foot-related problems. The main machine learning methods for medical diagnosis and rehabilitation procedures are analyzed. A systematic review and analysis of the use of machine learning methods in the context of plantar pressure analysis is conducted. Sample size, pathology, number of pressure sensors, data acquisition device, use of other sensor devices, preprocessing dataset, ML method varieties, and evaluation metrics are evaluated. Four machine learning methods are used for analysis and rehabilitation prediction: k-NN, Decision Tree, SVM, Logistic Regression. A comparative analysis of the methods' performance is performed, and the main metrics are obtained.