Detection of Parkinson’s Disease using Inceptionv3 and Exception with Residual Attention Network

Parkinson’s disease (PD) is a progressive neurodegenerative ailment that disturbs movement, cognitive function, and general value of life. Premature and precise findings are crucial for timely intervention and real administration. This study introduces a deep learning-based method that leverages InceptionV3 and Xception architectures, combined with a residual attention mechanism, to improve feature extraction and classification accuracy. The proposed method processes medical imaging data, emphasizing key patterns associated with PD through attention-driven feature refinement. The model undergoes training on the handwritten spiral dataset, leveraging transfer learning and fine-tuning techniques. Experimental results demonstrate improved accuracy and robustness in PD detection compared to conventional deep learning models, highlighting the effectiveness of residual attention in enhancing feature representation.