Publication

Abstract : Early detection of Alzheimer's is based on the evaluation of cognitive impairment based on magnetic resonance imaging. Our study presents a model based on the convolutional neural network (CNN) to classify cognitive impairment from magnetic resonance imaging. The model employs pre-processing techniques such as grayscale conversion, normalization, and image resizing to improve feature extraction. Its architecture consists of multiple convolutional and pooling layers, followed by fully connected layers with dropout regularization to mitigate overfitting. The model is trained and evaluated using categorical cross-entropy loss, with performance measured through accuracy and AUC-ROC/AUC-PR curves. The experimental results demonstrate the effectiveness of the model in classifying cognitive impairment, strengthening its potential for automated diagnosis based on magnetic resonance imaging