Comparative Analysis of Medical Diagnostic Techniques Using ANN

An immense and immeasurable amount of data is available to medical experts, extending from points of interest of clinical manifestations to different sorts of biochemical information and yields of imaging gadgets. Each kind of information yields data that must be assessed and relegated to a specific pathology amid the diagnostic process. To rationalize the diagnosis in every day routine and maintain a strategic distance from misdiagnosis, methods of machine learning (particularly ANNs) may be utilized. The versatile learning algorithms of machine learning may deal with several kinds of restorative heterogeneous information and classify them into various class outputs. In this paper, we concisely survey and examine the logic, capacities, and performance of ANNs in medical diagnosis of various diseases by making comparative analysis and focusing more on the medical diagnosis of Diabetes. The use of PID dataset for diagnosis is also demonstrated.