A Survey of Kidney Cancer Analysis Using Machine Learning and Deep Learning Algorithms

Abstract

A recent and emerging public health concern is kidney disease, particularly Chronic Kidney Disease (CKD), characterized by a gradual decline in kidney function due to various ailments. Kidney cancer, a potentially fatal form of cancer, requires accurate staging for effective treatment selection, with Stage 1 being a crucial point for decisions. Physicians often face challenges in accurately determining cancer stages, leading to potential under- or over-treatment. Kidneys play a vital role in filtering blood, removing waste and excess fluids, excreted as urine. Advanced renal disease can cause dangerous build-ups of fluids, electrolytes, and waste. In early stages, kidney disease may exhibit minimal symptoms, often progressing unnoticed until more advanced stages. Kidney cancer arises when kidney cells undergo abnormal growth and lose control. Treatments for kidney cancer encompass a range of options, including surgical procedures, chemotherapy, and radiation therapy. Early detection remains pivotal for effective intervention, echoing the universal importance seen in managing all types of cancer. In contrast, managing chronic kidney disease focuses on minimizing kidney damage by targeting its root causes, although addressing these factors may not always halt the disease's progression. In advanced stages, kidney cancer can potentially lead to kidney failure, a life-threatening condition necessitating artificial filtration methods like hemodialysis or surgical interventions for survival. “Various risk factors encompassing dietary habits, environmental exposures, and living conditions contribute significantly to sudden and often unnoticed illnesses among many individuals. Physicians commonly encounter challenges in accurately determining the stages of cancer, potentially leading to either under treatment or overtreatment. The diagnosis process for chronic renal disease is often fraught with risks, high costs, time-intensive procedures, and invasiveness. Consequently, numerous illnesses remain untreated until they reach advanced stages, especially prevalent in regions with limited resources, highlighting disparities in healthcare accessibility and early detection. Early detection is vital for doctors to save lives. Detecting kidney conditions early can help prevent kidney damage and slow down Kidney Cancer progression. “Researchers are exploring various methods to identify kidney disease at an early stage. This research aims to understand and improve early prediction techniques for this condition. To categorize Kidney cancer, researchers use different classification strategies like Artificial Neural Networks (ANN), Naive Bayes, Support Vector Machines (SVM), Adaptive Deep Convolutional Neural Network (AHDCNN), Alex Net, U-Net 2D, ada boost classifiers, and Linear Discriminant methods.