A Novel Framework for Human Chronic Disease Prediction Using CNN and Adaptive Dragon Fly Optimization (ADFO) Algorithm

Chronic diseases in human beings are long-lasting health disorders that will progress slowly over months or a lifetime. The chronic diseases can affect various parts or systems of the body and are non-acute in nature. The chronic disease does not occur suddenly and shall not be resolved quickly. However, the spreading of the disease to other parts of the body and the effectiveness of the disease shall be suppressed by providing proper medical treatment from the earlier stages. The World Health Organization has expressed the significance of the treatment of chronic disease, stating that around 41 million deaths, representing 71% of all deaths occurring every year due to chronic diseases. This necessitates the earlier detection of human chronic diseases like cancer, cardiovascular disease and diabetes mellitus. The major challenge associated with chronic disease is that it does not express valid symptoms at the earlier stage and needs advanced technology for the earlier detection process with high accuracy. The existing system of chronic disease predictions falls short in diagnosis and detection at the earlier stages, whereas capable of detecting at stage 3 or stage 4, which were considered the advanced stage of the chronic disease. To overcome this challenge, this proposed work introduces a novel framework encompassing the Convolutional Neural Networks (CNN) and the Adaptive Dragon Fly Optimization algorithm for the earlier prediction process. The proposed work was trained with datasets of various chronic diseases individually to prove the system’s adaptability. In addition, the performance of the proposed work is analyzed in terms of accuracy, Precision, recall and F score, which were compared with the performance of the existing state-of-the-art methodologies.