Application of 3D Convolutional Model Based on Artificial Intelligence in Parkinson’s Assessment

With the rapid development of artificial intelligence technology in the field of medical health, the Parkinson’s disease (PD) assessment method based on deep learning has become an important research direction. This paper proposes an artificial intelligence-based 3D convolutional neural network (3D-CNN) model for PD action assessment, aiming to minimize the cumulative errors common in traditional partitioning models. The model achieves an average accuracy of 79.25%, surpassing the existing state-of-the-art graph convolution-based method. We posit that our proposed framework has the potential to offer a more comprehensive and clinically acceptable approach for PD measurement, without imposing additional stress or workload on patients and healthcare professionals. This framework can also potentially be deployed for tracking PD progression in nursing homes, enabling self-assessment at home, and advancing telemedicine. It not only improves the level of automation in PD assessment but also has good prospects for clinical application, and can provide patients with more accurate and real-time health monitoring support.