Comprehensive Systematic Computation on Alzheimer's Disease Classification

Abstract

Alzheimer’s disease (AD) is a degenerative neurological ailment that progressively affects a large number of individuals globally. Timely and precise diagnosis of this ailment is crucial for effective therapy and control. In recent years, DL algorithms have demonstrated encouraging outcomes in assisting AD diagnosis by utilizing medical imaging datasets. Nevertheless, current DL models for AD classification encounter specific obstacles, including restricted interpretability and elevated computational cost. This article introduces a novel hybrid DL approach to address these difficulties. This model integrates conventional ML and DL techniques to perform better classification. The hybrid model is trained and tested using a substantial dataset comprising AD patients and individuals without AD. This study aims to comprehensively analyze the performance of the suggested hybrid model by comparing it to other advanced DL models already in use. The findings demonstrate that the proposed hybrid model surpasses current DL models in accuracy, sensitivity, and specificity. It possesses enhanced interpretability, facilitating doctors in effectively communicating the diagnosis to patients. The hybrid model exhibits reduced computing complexity, rendering it more efficient and feasible for real-time diagnosis. This study enhances the advancement of a novel hybrid DL model for AD classification by integrating the advantages of conventional ML algorithms and DL approaches. The results indicate the model's capacity for precise and efficient diagnosis. Subsequent studies can investigate the model's use with different medical imaging datasets to diagnose various neurodegenerative illnesses.