ML in Medical Imaging Informatics Recent Trends and Future Directions

The neurological condition known as Alzheimer's disease (AD) primarily impacting those over 60. At first, its symptoms are not severe, but they gradually worsen. There is no known cure for this illness, however early detection may reduce its effects. A technique For the prediction of AD, SMOTE has been proposed. Machine learning techniques are used to forecast Alzheimer's disease. The prediction performance of two techniques is assessed: Catboosting and Extreme gradient boosting (XGB). The study employs the Open Access Series of Imaging Studies (OASIS) longitudinal dataset, available on Kaggle, for experimentation. The dataset is balanced through the application of the synthetic minority oversampling technique. In the experiments, both balanced and unbalanced datasets are employed. CATBOOST achieved 76.91% accuracy, while XGB achieved 98.52%. In terms of best accuracy, SMOTE-XGB can reach up to 98.52%.