PARKINSON'S DISEASE EARLY DETECTION MODEL USING MACHINE LEARNING

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that primarily affects motor control and speech production due to the gradual degeneration of dopamine-producing neurons in the substantia nigra region of the brain. Early detection of Parkinson’s disease is essential for effective treatment and improved quality of life, yet traditional diagnostic methods rely heavily on clinical observations and neurological assessments that often detect the disease only in later stages. In recent years, machine learning techniques have shown significant potential in identifying early biomarkers associated with Parkinson’s disease, particularly through voice analysis. Speech impairment is one of the earliest symptoms observed in individuals with PD, making vocal biomarkers an important indicator for early diagnosis. This study proposes a machine learningbased Parkinson’s Disease Early Detection Model that utilizes vocal features extracted from patient voice recordings to identify the presence of the disease. Acoustic parameters such as jitter, shimmer, harmonic-to-noise ratio, fundamental frequency variations, and nonlinear dynamic measures are used as input features for classification algorithms. Various machine learning models including Support Vector Machine, Random Forest, and Logistic Regression are applied to classify healthy individuals and Parkinson’s patients based on extracted features. The proposed system focuses on improving diagnostic accuracy while maintaining a noninvasive and cost-effective approach. Experimental results demonstrate that machine learning models trained on voice datasets can effectively detect early signs of Parkinson’s disease with high accuracy. The study highlights the importance of artificial intelligence in medical diagnosis and demonstrates how speech signal processing combined with machine learning can provide an efficient, accessible, and scalable solution for early detection of Parkinson’s disease.