The Age of the Meta-Doctor: Diagnosing Parkinson’s Disease with Artificial Intelligence and Speech

Abstract

The basal ganglia consist of the striatum, substantia nigra, and other nuclei, forming various pathways of motor initiation. Parkinson’s disease (PD) is a neurodegenerative disorder characterized by dysfunction of the basal ganglia pathways. Consequently, PD affects the production of speech. An AI model can analyze audio samples from regular and PD patients. A simple deep learning model with various layers, ReLU activation, sigmoid activation, optimizer, loss function, and Early_Stopping can use extracted speech features to classify patients as regular or PD-afflicted with up to 97% accuracy. Overall, the advent of user-friendly artificial intelligence has led to exciting times, with new medical advancements emerging day after day; perhaps the ease of AI implementation will encourage others to solve everyday problems with just a computer and a dream.