A Machine Learning Framework for Accurate Prediction of Parkinson's Disease from Speech Data
Srinivasa Rao Burri, Deepak Agarwal, Narayan Vyas, Ronak Duggar
Abstract
Speech data can help Machine Learning (ML) systems diagnose Parkinson's disease (PD). This study predicted PD progression using the Parkinson's Progression Markers Initiative (PPMI) dataset and Deep Belief Networks (DBN). PD and healthy control voice inputs formed a Convolutional Neural Network (CNN) model. This DBN model was compared to others using numerous performance metrics. DBN outperformed other algorithms. It accurately and early detected PD. DBN-based models may enhance PD diagnosis. Complex neurodegeneration Parkinson's affects millions. Early diagnosis enhances medical therapy. Clinical evaluations and subjective judgments delay and misdiagnose traditional diagnostic methods. This study scientifically and data-driven diagnosed PD using the DBN model. Speech data can suggest physical and cognitive problems, which these algorithms can detect. This improves accuracy and permits early diagnosis before physical symptoms develop. This study suggests ML can transform PD diagnosis. PPMI-sized datasets help researchers increase model accuracy and reliability. Speech data lets doctors assess more. This research extends diagnosis. PD diagnosis and treatment improve disease management. Predicting PD progression using ML allows patient-specific treatment. This study indicates the DBN model can reliably and early diagnose Parkinson's disease using voice data. The DBN model shows ML can outperform other PD diagnosis approaches. These developments may enhance patient outcomes, illness management, and early interventions.