FUZZY-DRIVEN KIDNEY TUMOR DETECTION: INTEGRATING TWIN TRANSFERABLE NETWORKS WITH WEIGHTED ENSEMBLE MACHINE LEARNING

Abstract

Kidney cancer is one of the most prevalent urological malignancies worldwide, and its early detection plays a vital role in improving patient survival and treatment outcomes. Conventional diagnostic methods based on manual interpretation of Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scans are often time-consuming, subjective, and dependent on radiologists’ expertise, leading to diagnostic variability and delayed clinical decisions. Recent advances in Artificial Intelligence (AI), Deep Learning (DL), and Transfer Learning have significantly enhanced automated medical image analysis by enabling accurate tumor detection and classification. This paper proposes a fuzzy-driven kidney tumor detection framework that integrates Twin Transferable Networks (TTNs) with a Weighted Ensemble Machine Learning (WEML) model for intelligent medical diagnosis. The proposed framework combines fuzzy logicbased preprocessing, medical image enhancement, transfer learning, feature extraction, and weighted ensemble classification to accurately identify kidney tumors from CT images. Twin Transferable Networks automatically learn discriminative image representations, while the weighted ensemble model combines predictions from multiple classifiers to improve robustness and diagnostic accuracy. Experimental evaluation demonstrates that the proposed approach significantly outperforms conventional machine learning and standalone deep learning models in terms of accuracy, precision, recall, F1-score, and computational efficiency. The integration of fuzzy logic further enhances image quality by reducing uncertainty and improving feature discrimination. The proposed framework provides a reliable and scalable computer-aided diagnosis system capable of supporting radiologists in early kidney tumor detection, clinical decision-making, and precision healthcare applications.