Key points are not available for this paper at this time.
Impaired renal function poses a risk across all age groups. Because of the global shortage of nephrologists, the growing public health concern over renal failure, and technological improvements, there is a demand for an AI-driven system capable of autonomously detecting kidney abnormalities. Chronic kidney disease, commonly known as chronic renal failure, is characterized by a progressive decline in kidney function. Renal failure can be caused by a variety of reasons, including cysts, stones, and tumors. Chronic kidney disease may not have apparent symptoms at first, resulting in instances staying untreated until they reach an advanced state. Tumors are dense clumps of tissue that can cause direct injury to glands, spinal cells, and other organs. The presence of a substantial number of solids in the digestive tract causes kidney stone disease, also known as urolithiasis. This study used a deep learning model to detect kidney illnesses to solve the global scarcity of urologists. The project entailed obtaining and annotating a large dataset of 12,446 CT whole abdomen and urogram images, with an emphasis on kidney stones, cysts, and tumors, which are the most common types of renal illness. The dataset was divided into four categories: cyst, tumor, stone, and normal. Data was collected from several hospitals in the Dhaka area. This work presents an innovative and customizable platform for the clinical diagnosis of kidney diseases such as tumors, stones, and cysts. Our YOLOv8 model’s enhanced accuracy opens up new possibilities for identifying kidney cysts, stones, and tumors. We used criteria like accuracy, precision, recall, F1 score, and specificity to evaluate its performance. The network attained an accuracy rate of 82.52%, 85.76% precision, 75.28% recall, 75.72% F1 score, and 93.12% specificity.
Pande et al. (Mon,) studied this question.