A Detection and Prediction Model Based on Deep Learning Assisted by Explainable Artificial Intelligence for Kidney Diseases

Authors : Ahmet Furkan BAYRAM, Caglar GURKAN, Abdulkadir BUDAK, Hakan KARATAŞ

Pages : 67-74

Doi:10.31590/ejosat.1171777

View : 22 | Download : 12

Publication Date : 2022-09-30

Article Type : Research Paper

Abstract :Kidney diseases are one of the most common diseases worldwide and cause unbearable pain in most people. In this study aims to detecting the cyst and stone in the kidney. For the this purpose, YOLO architecture designs were used for detection of kidney, kidney cyst and kidney stone. The YOLO architecture designs were supported by the explainable artificial intelligence insert ignore into journalissuearticles values(xAI); feature. CT images in three classes, namely 72 kidney cysts, 394 kidney stones and 192 healthy kidneys were used in the performance analysis part of the YOLO architecture designs. As a result, YOLOv7 architecture design outperformed the YOLOv7 Tiny architecture design. YOLOv7 architecture design achieved the mAP50 of 0.85, precision of 0.882, sensitivity of 0.829 and F1 score of 0.854. Consequently, deep learning based xAI assisted computer aided diagnosis insert ignore into journalissuearticles values(CAD); system was developed for diagnosis of kidney diseases.
Keywords : Böbrek taşı, Böbrek kisti, Derin öğrenme, YOLOv7, Açıklanabilir yapay zeka