Enhancing Parkinson's Disorder Detection Accuracy: A Machine Learning Approach for Analyzing Voice Signals

Parkinson's disease is an autoimmune disorder characterized by the death of neurons in the brain that synthesize dopamine, which results in a broad spectrum of symptoms. Among these, speech and voice-related issues, collectively known as Parkinson's-related voice or speech disorders, are common and significantly affect patients' communication abilities. Resting tremors, a hallmark symptom of Parkinson's, can affect the vocal cords, causing a shaky or trembling voice. Patients often exhibit reduced volume, speaking softly and making it challenging for others to hear them. Additionally, some individuals may experience stuttering or stammering as a result of the disease's impact on speech control. To address the challenge of timely identification of Parkinson's disease, it has been brought forward to use a multilayer neural network that utilizes machine learning models and a curated dataset by Max Little. This system employs meticulous feature selection of the voice signals and data preprocessing techniques to enhance model performance. Among the various machine learning methods explored, the Multilayer Neural Network stands out, achieving exceptional training and testing accuracy rates of 100.00% and 98.43 %, particularly excelling in distinguishing between positive and negative cases. The effectiveness of this method demonstrates the capability of machine learning to enhance the initial identification and therapy of Parkinson's disease. It offers a powerful tool for healthcare professionals and researchers, aiding in patient care and advancing our understanding of this complex condition. By leveraging data-driven insights, this system aims to make significant strides in the realm of Parkinson's disorder. At some point, this will improve patient outcomes and quality of life.