ENHANCED ALZHEIMER DISEASE DIAGNOSIS THROUGH DEEP LEARNING-BASED MRI SEGMENTATION AND CLASSIFICATION
: https://doi.org/10.65725/JCISE/2/3/002
JOURNAL OF COMPUTATIONAL INTELLIGENCE SCIENCE AND ENGINEERING (JCISE)
ISSN: 3107-8168
Volume 2 Issue 3, Jul-Sep 2026
Abstract:
The accurate diagnosis tends to remain critical for the improvement of the patient outcomes in the Alzheimer’s disease (AD), where a progressive neurodegenerative disorder with the growing of the global impact. MRI has become an effective modality for analyzing the brain structure, which offers a detailed insights into the pathological regions that is associated with the AD. Deep learning has applied to the MRI data that presents a promising direction because it can support a higher diagnostic precision. However, the diagnosis of AD in its early stages tends to remain challenging due to the slow disease progression and the overlapping symptoms. This often causes a subtle structural change in the brain to be overlooked during the conventional assessment. The complexity of the MRI interpretation and the need to precisely identify the degenerative tissue indicates the importance of an advanced image segmentation strategy. This study have addressed these challenges by the combination of deep learning with the automated segmentation for the early AD detection. The proposed survey have exploited the MRI scans to analyze the structural changes in the brain, with a primary focus on the hippocampal extraction as a biomarker. The approach uses the OASIS-2 that gets trimmed with the MRI dataset for a rapid and an efficient training and this applies the Kaggle AD the MRI dataset for severity classification. The performance is then evaluated using the segmentation accuracy and the other model performance metrics. The method has also shown the improved diagnostic capability when it is compared with the existing approaches.
Authors: Dr K S Thirunavukkarasu, J.Johnsirani
Keywords: Alzheimer’s disease, the MRI, hippocampus segmentation, deep learning, classification.