#High-Accuracy Detection of Early Parkinson's Disease through Multimodal Features and Machine Learning Personal Summary

This paper is about early detection of Parkinson's and how researchers use the help of biomarkers (like CSF), dopaminergic transporter quantity, and pre motor symptoms like RBD and hyposmia. Using features of patients from conclusive tests and features (UPSIT, SS, biomarkers), it compares the specificity, sensitivity, and accuracy of different classifiers to determine what classifier is best for early detection of PD. From the tests and statistical analysis, it was found that Support Vector Machines (SVM) are the most accurate and best classifiers to use in this given situation. The research performed in this paper is considered an extension of others with more accuracy with the use of more subjects. It hopes that the SVM classifier can be used by clinicians to detect PD early before it does significant neurological damage.