A Review of the Literature on Automated Parkinson's Disease Diagnosis Methods Using Machine Learning

Abstract

Clinical signs, such as the description of a variety of movement symptoms and medical observations, are frequently used to diagnose Parkinson's disease (PD). Conventional diagnostic techniques may be subjectivity-prone since they rely on the interpretation of motions that may be challenging to identify because they are occasionally imperceptible to the human sight. Meanwhile, early Parkinson's disease non-motor symptoms may be slight and brought on with a variety of other illnesses. Therefore, it might be difficult to diagnose this disease in the early stages because the symptoms are frequently disregarded. To classify this disease, machine learning methods have been created. This chapter includes a review of the literature for works released through 2023 utilizing the number of databases in order to give a thorough outline of machine learning methods used in the identification and classification of PD.