A Survey of Predicting CKD Using Machine Learning

Abstract

Chronic kidney disease (CKD) poses a significant global public health challenge, affecting approximately 10% of the worldwide population. Despite recent increases in awareness, understanding of the disease remains limited. Alarmingly, the incidence, morbidity, mortality, and associated healthcare costs of CKD continue to rise, especially in low-income countries.  Chronic kidney disease (CKD) represents the most severe stage of kidney damage, where the kidneys gradually lose functionality and may eventually cease to function entirely. Key risk factors for CKD include high blood pressure, cardiovascular disease, diabetes, advanced age, and a family history of kidney failure. Secondary risk factors encompass obesity, autoimmune diseases, systemic infections, urinary tract infections, and other kidney-related issues such as kidney damage, injury, or infection. Treatment strategies for CKD vary based on the patient's physical condition and typically involve lifestyle modifications, medications to manage related health problems, dialysis, and ultimately, kidney transplantation.  Early diagnosis is crucial for effective treatment of CKD. The two primary methods for diagnosing CKD are blood and urine tests. However, these manual processes require expert involvement, which can be time-consuming and resource-intensive. To address these challenges, recent research has focused on developing automated, computerized diagnostic approaches using artificial intelligence. In this context, machine learning (ML) has emerged as the preferred choice among researchers due to its efficiency and accuracy.