Image Recognition of Alzheimer's Disease Using Machine Learning

Alzheimer's disease is the most common type of dementia, which is not easily detected in its early stage of development and has become one of the killers endangering human health. Traditional machine learning models exhibit limited accuracy on small-to-medium medical datasets, while deep learning requires extensive data and computational resources. This study performs multi-CNN, which can capture deep features and compare other machine/deep learning models, including Random Forest (RF), Support Vector Machine (SVM), Multilayer Perceptron (MLP), and ResNet. Experiments were conducted on the OASIS MRI dataset. The results indicate that the accuracy of the multi-CNN achieved an accuracy of 0.90, performing the best in indicators including metrics such as precision, recall, and the F1 score, which is significantly higher than other machine learning models. These findings highlight deep learning models' robustness in medium-scale medical imaging, offering a cost-effective AI tool for early AD detection in resource-constrained clinical settings.