Parkinson Disease Detection by Analyzing Spiral Drawings

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by motor symptoms such as tremors, bradykinesia, and rigidity. Early and accurate diagnosis of PD is crucial for effective treatment and management. This paper presents a novel approach for PD detection using spiral drawings and machine learning techniques. We utilized a dataset comprising spiral drawings collected from individuals with PD and healthy controls. Image preprocessing techniques were applied to enhance image quality and standardize format. Features capturing motor abnormalities were extracted from the spiral images, including measures of symmetry, curvature, and tremor frequency. A random forest regression model was trained on the extracted features to predict the severity of motor impairment associated with PD. Cross-validation techniques were employed to assess model performance, with metrics such as mean squared error and receiver operating characteristic analysis used for evaluation. The trained model demonstrated promising results in distinguishing between individuals with PD and healthy controls. Interpretability of the model was enhanced through visualization techniques, providing insights into the underlying patterns contributing to PD diagnosis. Our approach offers a non-invasive and cost-effective tool for early PD detection, with potential applications in clinical practice for improving patient outcomes. This abstract summarizes the key components and findings of the paper, highlighting the novelty and potential impact of the proposed methodology for PD detection using spiral drawings and machine learning.