Artificial Intelligence for Iris-Based Diagnosis in Healthcare
Computers are now able to play a role in the diagnostic process for diabetes patients because of developments in the disciplines of medical imaging and information technology. It is difficult to reduce death rates to more reasonable levels by employing early and precise detection approaches. When seeking to identify risk factors for complications, acute and long-term difficulties, mortality, and medical care expenses, data on the prevalence and incidence of diabetes and prediabetes is important. The International Diabetes Federation estimates the occurrence of diabetes based on the present status and prevalence of diabetes worldwide. In addition, variables, such as physiological markers, that may have a role in the development of diabetes are investigated. Complementary and alternative methods, especially iris-based diagnostics, have made significant advances in the identification of diabetes and associated disorders. Diabetes is a leading cause of renal failure globally. It raises the likelihood of heart and blood vessel problems. Diabetic kidney disease is a chronic condition. The blood test is used to identify diabetic kidney disease, which is a difficult and agitating treatment for patients. This chapter combines iridology with current computer-based approaches to diagnose diabetic kidney disease. A total of 130 participants were assessed and categorized based on their health status, such as not having diabetes, having diabetes but not having diabetic kidney disease, and having diabetic kidney disease. Pre-image processing methods are used to extract the region of interest from the iris based on the iridology chart’s specified regions. The region of interest was analyzed for first-order statistical and second-order textual characteristics. Finally, several categorization models were examined using different sets of characteristics to discover the optimum model. The maximum accuracy of 94.4% was achieved with the help of a fine Gaussian kernel of the SVM classifier. It is empirically demonstrated that features of specific areas of the iris are highly correlated with the diabetic kidney disease condition of an individual.
Ravinder Agarwal, Piyush Samant, Atul Bansal & Rohit Agarwal