Answers are grounded in this paper and cite their sources.
Alzheimer's disease (AD) is a progressive form of neurodegenerative disorder causing impairments in memory functions and cognitive abilities. Therefore, early diagnosis of AD or Mild Cognitive Impairment (MCI) is an essential step toward the successful treatment of patients. Many of the existing classifiers for Alzheimer's disease depend heavily on utilizing deep learning algorithms along with MRI image processing to identify patterns related to the dementia disorder. Nevertheless, many state-of-the-art methods primarily depend on imaging-derived features and do not effectively incorporate neurocognitive accuracy, relying predominantly on imaging-based features. This study proposes a Neuro Cognitive Adaptive Spectral Recalibration Convolutional Neural Network (NCASR-CNN) Framework as a technique for detecting Alzheimer's disease with the use of OASIS MRI datasets. The proposed NCASR-CNN model incorporates the clinical variables age and Mini-Mental State Examination (MMSE) to enhance the discriminative capability of extracted features. Final classification was performed using a hybrid stacking ensemble comprising a Multi-Layer Perceptron (MLP), LightGBM, XGBoost, and Logistic Regression, using three diagnostic categories: Alzheimer's disease, Mild Cognitive Impairment, and Cognitively Normal (CN) subjects. The proposed framework achieved an overall ROC–AUC of 99.70% and a classification accuracy of 94.38%, while also attaining high precision, recall, and F1-score across all diagnostic categories. The results show that integration of Reinforcement Learning (RL) to choose slices, adaptive parameter optimisation, combining deep features, and hybrid ensemble learning work well-together for accurately classifying Alzheimer's disease.
No takes yet. Share an insight, caveat, or question.
Bhaskar et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: