Deep Learning Model-based Decision Support System for Kidney Cancer on Renal Images

• Published in: Engineering, technology & applied science research Vol. 14; no. 5; pp. 17177 - 17187
• Main Authors: Tounsi, Mohamed, Abdulhussain, Donya Y., Azar, Ahmad Taher, Al-Khayyat, Ahmed, Ibraheem, Ibraheem Kasim
• Format: Journal Article
• Language: English
• Published: 09.10.2024
• ISSN: 2241-4487; 1792-8036
• DOI: 10.48084/etasr.8335

Kidney cancer comes in various forms. Renal Cell Carcinoma (RCC) is the most severe and common kind of kidney cancer. Earlier diagnosis of kidney cancer has enormous advantages in implementing preventive measures to reduce its effects and death rates and overcome the tumor. Manually detecting Whole Slide Images (WSI) of renal tissues is a basic approach to predicting and diagnosing RCC. However, manual analysis of RCC is prone to inter-subject variability and is time-consuming. Compared to time-consuming and tedious classical diagnostic methods, automatic Deep Learning (DL) detection algorithms can improve test accuracy and reduce diagnostic time, radiologist workload, and costs. The study presents a Computational Intelligence with a Deep Learning Decision Support System for Kidney Cancer (CIDL-DSSKC) technique on renal images. The CIDL-DSSKC model analyzes renal images to identify and classify kidney cancer. The proposed method uses Median and Wiener filters for image preprocessing and the Xception model to derive a useful set of feature vectors. In addition, the Flower Pollination Algorithm (FPA) is employed to optimally choose parameters for the Xception method. The