ADC-RBK: a multimodal approach for early detection of chronic diseases and focusing on Alzheimer’s

Chronic diseases are the most deadly and complex diseases to be treated across the world. It affects the daily activities of the patients and makes them bedridden. Millions of individuals are affected by chronic diseases annually. Some of the chronic diseases are arthritis, Alzheimer’s, asthma, cancer, diabetes, heart disease, etc. Early detection of these diseases can prevent the death of patients and can increase the lifespan. Even though there are various diagnostic measures for the detection of these diseases in earlier stages they have some limitations. Hence there is a requirement for an efficient automatic multimodal disease risk prediction system related to chronic diseases. Therefore, this paper developed an Attention Dual Convolutional-based Random Binary Kepler (ADC-RBK) algorithm. In this method, the images are preprocessed initially which are obtained from the Alzherimer’s diseases dataset. This dataset includes MRI images of the brain in order to discover Alzheimer’s disease. Subsequently, the preprocessed data is entered into the risk prediction phase. The risk prediction is performed with the ADC-RBK method. A Dual CNN assisted in overcoming the challenges of concentrating the local features by the CNN, thus it learns the features from the multiple domains. The feature extraction process is enhanced with the Attention mechanisms by focusing on the specific regions of the images and finally, the Random Binary Kepler algorithm optimizes the classification method to further enhance the model. Various evaluation metrics namely accuracy, precision, recall, and F1-score are utilized to evaluate the performance of the ADC-RBK method and compare its performance with existing methods. Experimental outcomes performed with certain measures and comparisons with existing methodologies show the model has promising outcomes with 98.1% accuracy, 97.9% precision, and 98.6% recall.