Towards Personalized Medicine: Machine Learning for Parkinson’s Disease Diagnosis

Parkinson’s disease (PD) is a neurodegenerative disorder where early diagnosis is crucial for effective management. This study explores the potential of voice-based biomarkers, utilizing speech recordings, to forecast PD onset or progression through machine-learning techniques. The dataset incorporates speech features from PD patients and healthy individuals, analyzed using various algorithms including Support Vector Machine (SVM), Random Forest, Logistic Regression, and Neural Networks. Preprocessing methods such as feature scaling and train-test partitioning are applied. Model effectiveness is evaluated using metrics like F1-score, recall, accuracy, and precision, with feature importance analysis identifying crucial voice characteristics for PD prediction. Results demonstrate a clear differentiation between PD patients and healthy controls, with SVM exhibiting the highest accuracy. These findings underscore the promise of voice-based biomarkers for early PD diagnosis and advancements in non-invasive diagnostics. Future research will focus on validating models in clinical settings and exploring larger datasets for improved predictive accuracy and clinical applicability.