Detection of neurodegenerativediseases by the automaticanalysis of handwriting

Abstract

Neurodegenerativediseases (ND) are a serious issue whichencompasses a myriad of complex and incurable disorders. in thisthesis, we focus on Parkinson’sdisease (PD), specifically; the detection of PD though the automaticanalysis of offline handwriting. To accomplishthistask; we propose Park-Net, ourownconvolutional neural network (CNN) architecture. Weproceed to test this CNN on three PD handwritingdatasetsbeforecomparing the results to state-of-the-art works, and with a 98%accuracy, and to the best of ourknowledge; Park-Net outperformsstudies as recent as (2022).