PROJECT TITLE :

Computer-Aided Diagnosis of Chronic Kidney Disease in Developing Countries A Comparative Analysis of Machine Learning Techniques

ABSTRACT:

The high incidence and prevalence of chronic kidney disease (CKD), which is sometimes caused by late diagnosis, is a major public health issue, particularly in developing nations like Brazil. Dialysis and kidney transplantation are examples of CKD treatment treatments that raise morbidity and mortality rates while also increasing public health expenses. The use of machine learning techniques to aid in the early detection of CKD in underdeveloped nations is examined in this study. A systematic literature review and an experiment with machine learning approaches, employing the k-fold cross-validation method based on the Weka software and a CKD dataset, are used to undertake qualitative and quantitative comparative evaluations, respectively. These findings enable a discussion on the applicability of machine learning techniques for CKD risk screening, with a focus on low-income and hard-to-reach settings in developing nations, due to the unique challenges they confront, such as poor primary health care. The study findings show that the J48 decision tree is a suitable machine learning technique for such screening in developing countries, owing to the ease with which its classification results can be interpreted, with 95.00 percent accuracy and nearly perfect agreement with the opinion of an experienced nephrologist. Random forest, naive Bayes, support vector machine, multilayer perceptron, and k-nearest neighbor techniques, on the other hand, produce 93.33 percent, 88.33 percent, 76.66 percent, 75.00 percent, and 71.67 percent accuracy, respectively, indicating at least moderate agreement with the nephrologist at the cost of a more difficult interpretation of the classification results.