Exploring Advanced Diagnostic Approaches for Parkinson’s Disease to Enhance Health, Well-Being, and Resilience in Sustainable Societies

This study aims to improve early Parkinson’s disease (PD) diagnosis through handwriting analysis, focusing on motor symptoms that often appear before formal diagnosis, thus facilitating better healthcare strategies. By integrating Deep Learning (DL) and coordinate attention scheduling (CAS) transformers, this approach addresses the challenges of recognizing subtle variations in handwriting for PD detection, which existing methods fail to fully capture. The proposed deep learning model utilizes CAS transformers to enhance handwriting feature extraction, analyzed across two handwritten datasets related to Parkinson's disease symptoms. The model demonstrated a classification accuracy of ۹۲.۶۸%, outperforming traditional methods in identifying critical features related to PD, highlighting its potential in early-stage diagnosis. This approach can improve early diagnosis and intervention, contributing to more sustainable, resilient healthcare systems. Its practical application could revolutionize the management of neurodegenerative diseases. By leveraging CAS transformers, this study offers a novel solution with superior diagnostic accuracy, contributing significantly to both scientific research and sustainable healthcare practice.

کلیدواژه ها:

CAS Transformer ، Deep Learning ، Early Diagnosis ، Handwriting Analysis ، Neurological Health ، Parkinson’s Disease