Predicting Parkinson’s Disease Using Machine Learning with Voice Parameters and Handwriting Images

Abstract

Most studies have failed to focus on geriatric diseases in the present era of quick advancement in medical science. Diseases like Parkinson’s display their symptoms at a later stage and make a complete recovery almost doubtful. Parkinson’s disease is a neurodegenerative disorder that affects movement and motor control systems. It is named after Dr. James Parkinson, the first person affected by this disease. Parkinson’s slowly worsens over time, leading to a variety of syndromes that can impact a person’s daily life activities. More than 95% of Parkinson’s Disease (PD) patients stated that they have exhibited voice impairment and micrographic disability. This model takes advantage of both advanced machine learning algorithms and modern image processing techniques, resulting in effective and efficient prediction PD. To further enhance the accuracy of the model, we have incorporated additional algorithms such as Random Forest and K-nearest Neighbour. Random forest classifier has a detection accuracy of 92%and sensitivity of 0.95%. The performance has been assessed with a reliable dataset from the University of California Irvine Machine Learning repository for voice parameters and a dataset from Kaggle for Handwriting images which includes wavy images and spiral images. Our proposed model has achieved the highest accuracy of 95% which outperformed the previous model or experiment on the same dataset.