Alzheimer’s Disease (AD), which affects the elderly population worldwide, is an acute degenerative disease. Hence, to treat the disease, early diagnosis is required. For obtaining an accurate diagnosis, various methods have been developed by the prevailing research. However, improvement is still needed. Therefore, this article proposes a Neglect VGGNet (NV) classifier detection utilizing a Magnetic Resonance Imaging (MRI) image. Primarily, the input MRI images are taken from different dimensions. By using Edge Quality Indexed Lee Filter (EQILF), Contrast-Limited Adaptive Histogram Equalization (CLAHE), and the morphological operation process, all the dimensions are pre-processed. Thereafter, the axial view is given to the segmentation process for segmenting the brain tissues utilizing the Frechet K-Means (FKM) algorithm. From the segmentation outcome, the hippocampus and neocortex have similar structures. Thus, to obtain accurate AD detection, separation is needed. Therefore, to find the difference, this research methodology uses Bellman Weighted Temporal Difference Learning (BWTDL). Afterward, the features are extracted from all the brain tissues; then, by utilizing the Sin Functional Lyrebird Optimization Algorithm (SFLOA), the important features are selected. By using a Point by Point Chord-based Structural Relational Graph (PPC-SRG), the graph is constructed from the sagittal view, and the matrix is generated. Moreover, by using the Canny Edge Aware Pyramidal Deformable (CEAPD) approach, the mapping is constructed from the coronal view. Thereafter, for predicting the AD, the selected features, generated matrix, and constructed map function are given to the NV classifier. The Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset is utilized for the performance analysis. As per the experimental evaluation, the proposed methodology attains 98.6% accuracy, which is higher in contrast to the prevailing mechanisms.
Dwarkanth et al. (Wed,) studied this question.