KDH-Net: Explainable Medical AI for Multiclass Kidney Disease Characterization from CT Images
Journal of Clinical Medicine · 21.04.2026 · 15 (8) · 3165
DOI: 10.3390/jcm15083165 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Clinical Medicine, online erschienen am 21.04.2026. Verfasst von Md Serajun Nabi und 9 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of AI and Engineering, Multimedia University, Persiaran Multimedia, Cyberjava 63100, Malaysia. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Accurate differentiation of kidney diseases such as cysts, tumors, stones, and normal tissue from computed tomography (CT) images remains challenging due to overlapping visual characteristics and variability in data distributions. While deep learning approaches have shown promising results, many existing studies rely on image-level data splitting and focus primarily on accuracy, which may lead to overly optimistic performance and limited clinical reliability.
Methods
This study proposes KDH-Net (Kidney Disease Hybrid Network), a hybrid deep learning framework for multiclass kidney disease characterization that integrates EfficientNetB0, ResNet50, and MobileNetV2 through feature-level fusion. A two-stage training strategy is adopted to enhance optimization stability. To ensure realistic performance assessment, experiments on the primary dataset are conducted under a patient-level evaluation protocol, eliminating potential data leakage. The framework further incorporates calibration analysis, statistical validation, and explainable artificial intelligence to evaluate prediction reliability and interpretability.
Results
On the patient-level dataset, KDH-Net achieves an overall accuracy of 0.93 with a macro-average F1-score of 0.91, demonstrating balanced performance across all classes. Confidence analysis indicates meaningful alignment between prediction confidence and correctness, while Grad-CAM visualizations highlight anatomically relevant regions associated with each class.
Conclusions
The results demonstrate that KDH-Net provides a stable, reliable, and interpretable framework for kidney CT characterization. The proposed system is designed to support clinical decision-making by offering trustworthy predictions under realistic evaluation conditions, rather than replacing clinical expertise.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Faculty of AI and Engineering, Multimedia University, Persiaran Multimedia, Cyberjava 63100, Malaysia.
- School of Computing and Informatics, Albukhary International University, Alor Setar 05200, Kedah, Malaysia.
- Faculty of Information Science and Technology, Multimedia University, Jalan Ayer Keroh Lama, Bukit Beruang 75450, Malaysia.
- Centre for Image and Vision Computing, COE for Artificial Intelligence, Multimedia University, Cyberjaya 63100, Malaysia.
ORCID-Profile: Md Serajun Nabi, Hasanul Bannah, Arbab Sufyan Wadood, Hezerul Bin Abdul Karim
Zitieren
APA
Nabi, M. S., Tun, S. W., Alam, S., Abdullahi, M. K., Bannah, H., Santo, I. A., Wadood, A. S., Md Mohiuddin, G., Rehman, Z. U., & Karim, H. B. A. (2026). KDH-Net: Explainable Medical AI for Multiclass Kidney Disease Characterization from CT Images. Journal of Clinical Medicine, 15(8), 3165. https://doi.org/10.3390/jcm15083165
Vancouver
Nabi MS, Tun SW, Alam S, Abdullahi MK, Bannah H, Santo IA, et al. KDH-Net: Explainable Medical AI for Multiclass Kidney Disease Characterization from CT Images. Journal of Clinical Medicine. 2026;15(8):3165. doi:10.3390/jcm15083165
BibTeX
@article{Nabi2026,
author = {Nabi, Md Serajun and Tun, Su Waddy and Alam, Shahaba and Abdullahi, Muhammad Kabir and Bannah, Hasanul and Santo, Istiyak Amin and Wadood, Arbab Sufyan and Md Mohiuddin, Golam and Rehman, Zaka Ur and Karim, Hezerul Bin Abdul},
title = {KDH-Net: Explainable Medical AI for Multiclass Kidney Disease Characterization from CT Images},
journal = {Journal of Clinical Medicine},
year = {2026},
volume = {15},
number = {8},
pages = {3165},
doi = {10.3390/jcm15083165}
}
EUVIMED