DLFM: Leveraging Parkinson's Disease Detection using AI Deep Learning Fusion Model for Precise Diagnosis

Parkinson's disease (PD) presents substantial difficulties owing to its progressive course and heterogeneous symptoms, thus an accurate and timely diagnosis is vital for optimal management. In this paper, we introduce a mixed AI Deep Learning Fusion Model (DLFM) as a new approach to improve PD detection. The DLFM combines the features obtained from both biomedical voice measurements and clinical examinations by utilizing the LeNet-5 and DenseNet architectures; thereby providing a robust mechanism for ensuring maximum diagnostic accuracy. The approach of our approach in regard to balancing and the process are as follows, to train the DLFM model with appropriate PD case classification we pre-process the dataset and extract some key features then, the DLFM paradigm that incorporates DL and fusion techniques develops a strong basis for precise PD diagnosis, enabling early intervention and personalized treatments. It shows better performance than conventional diagnostic solutions because it combines both LeNet-5 and DenseNet architectures, which give the ability to detect complex patterns and correlations between input data. Additionally, the DLFM seamlessly integrates information from multiple sources together and provides a robust diagnosis of PD status. Our results highlight the promise of AI-driven approaches to transform PD patient diagnosis and care. With a model accuracy of 97.23%, it demonstrates an excellent capability for distinguishing between PD and healthy patients. Adopting this DLFM model with high accuracy will provide medical doctors and researchers with a useful complementary tool to assist in earlier diagnosis of PD and timely therapeutic action. The DLFM model rapidly advances the path to an improved method for diagnosing PD through the use of state-of-the-art DL methodologies, resulting in improved patient outcomes and quality of life.