International Journal of Engineering and Manufacturing (IJEM)
IJEM Vol. 16, No. 5, 8 Oct. 2026
Cover page and Table of Contents: PDF (size: 1086KB)
Multi-Resolution Attention-Fused, Calibration-Aware Diagnosis of Kidney CT (Normal/Cyst/Stone/Tumour)
PDF (1086KB), PP.255-271
Views: 0 Downloads: 0
Author(s)
Shanker M. C.
1
N. Sankar Ram
2
V. Gokula Krishnan
3,*
Pinagadi Venkateswararao
4
Therasa Michael
5
S. Kaviarasan
6
Index Terms
Computed tomography, Multi-class Brier term, Tiny high-contrast targets, Calibration-aware objective, Class-balanced sampling, Normalization
Abstract
Accurate classification of kidney abnormalities from computed tomography (CT) images is essential for early diagnosis and clinical decision-making. However, distinguishing kidney stones, cysts, tumours, and normal kidneys remains challenging because of variations in lesion size, appearance, and imaging conditions. This study proposes a multi-resolution attention-fused, calibration-aware deep learning framework for four-class kidney CT image classification. The framework integrates convolutional neural networks and a Vision Transformer to extract complementary local and global features, which are combined through an adaptive attention-based fusion mechanism. Calibration-aware learning, incorporating focal loss, multi-class Brier loss, and feature-level orthogonality regularization, is employed to improve prediction reliability and confidence estimation. The model was trained and evaluated using an image-level data split on a publicly available kidney CT dataset containing normal, cyst, stone, and tumour images. Experimental results demonstrate an overall classification accuracy of 99.3%, a Macro-F1 score of 99.1%, an AUROC of 0.999, and an AUPRC of 0.998 on the independent test set. Robustness analysis under synthetic image corruptions showed only a modest reduction in performance, with the AUROC remaining at 0.995 under severe motion blur conditions. External validation on two independent cohorts (External Site-A and External Site-B) achieved AUROC values of 0.993 and 0.988, respectively, while post-hoc temperature scaling improved calibration by reducing the Expected Calibration Error (ECE) from 0.028 to 0.012 and 0.034 to 0.015. These results demonstrate that the proposed framework provides accurate, robust, and well-calibrated predictions, supporting its potential application in computer-aided kidney CT image analysis.
Cite This Paper
Shanker M. C., N. Sankar Ram, V. Gokula Krishnan, Pinagadi Venkateswararao, Therasa Michael, S. Kaviarasan, "Multi-Resolution Attention-Fused, Calibration-Aware Diagnosis of Kidney CT (Normal/Cyst/Stone/Tumour)", International Journal of Engineering and Manufacturing (IJEM), Vol.16, No.5, pp. 255-271, 2026. DOI:10.5815/ijem.2026.05.14
Reference
[1]Islam, M.S.B., Sumon, M.S.I., Sarmun, R. and Bhuiyan, E.H. (2024). Classification and segmentation of kidney MRI images for chronic kidney disease detection. *Computers and Electrical Engineering*. 119; 109613.
[2]Almuayqil, S.N., Abd El-Ghany, S., Abd El-Aziz, A.A. and Elmogy, M. (2024). KidneyNet: A novel CNN-based technique for the automated diagnosis of chronic kidney diseases from CT scans. *Electronics*. 13(24); 4981.
[3]Maniyar, M.F., Nagtode, Y., Patil, B. and Patil, G. (2024). Kidney disease classification using convolutional neural networks (CNN) on CT scan images. *Kidney*. 4(3).
[4]Sasikaladevi, N. and Revathi, A. (2024). Digital twin of renal system with CT-radiography for the early diagnosis of chronic kidney diseases. *Biomedical Signal Processing and Control*. 88; 105632.
[5]Stephe, S., Manjunatha, B., Revathi, V. et al. (2025). Osteosarcoma cancer detection using Ghost-Faster RCNN model from histopathological images. *Iran Journal of Computer Science*. 8; 217–231.
[6]Pande, S.D. and Agarwal, R. (2024). Multi-class kidney abnormalities detecting novel system through computed tomography. *IEEE Access*. 12; 21147–21155.
[7]Luo, H., Li, J., Huang, H., Jiao, L., Zheng, S., Ying, Y. and Li, Q. (2024). AI-based segmentation of renal enhanced CT images for quantitative evaluation of chronic kidney disease. *Scientific Reports*. 14(1); 16890.
[8]Ganie, S.M., Pramanik, P.K.D. and Zhao, Z. (2024). Deep transfer learning for kidney disease detection using CT scan images. In: *Proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM)*; 5624–5629.
[9]Zhang, M., Ye, Z., Yuan, E., Lv, X., Zhang, Y., Tan, Y. et al. (2024). Imaging-based deep learning in kidney diseases: Recent progress and future prospects. *Insights into Imaging*. 15(1); 50.
[10]Kumar, Y., Brar, T.P.S., Kaur, C. and Singh, C. (2024). A comprehensive study of deep learning methods for kidney tumour, cyst, and stone diagnostics and detection using CT images. *Archives of Computational Methods in Engineering*. 31(7); 4163–4188.
[11]Nazmul0087. CT Kidney Dataset: Normal, Cyst, Tumor and Stone. Kaggle. Available at: https://www.kaggle.com/datasets/nazmul0087/ct-kidney-dataset-normal-cyst-tumor-and-stone
[12]Rana, S.U., Nur-A-Alam, M., Akter, S. and Likhon, M.N.H. (2025). KidneyMultiNet: A web-based automatic system for kidney disease detection using hybrid machine learning model from CT scan images. *Biomedical Materials & Devices*. 3(2); 885–896.
[13]Pimpalkar, A., Saini, D.K.J.B., Shelke, N., Balodi, A., Rapate, G. and Tolani, M. (2025). Fine-tuned deep learning models for early detection and classification of kidney conditions in CT imaging. *Scientific Reports*. 15(1); 10741.
[14]Ayogu, I.I., Daniel, C.F., Ayogu, B.A., Odii, J.N., Okpalla, C.L. and Nwokorie, E.C. (2025). Investigation of ensembles of deep learning models for improved chronic kidney disease detection in CT scan images. *Franklin Open*. 100298.
[15]Sheng, T.W., Onthoni, D.D., Gupta, P., Lee, T.H. and Sahoo, P.K. (2025). Segmentation of ADPKD computed tomography images with deep learning approach for predicting total kidney volume. *Biomedicines*. 13(2); 263.
[16]Sharma, K., Uddin, Z., Wadal, A. and Gupta, D. (2025). Hybrid deep learning framework for classification of kidney CT images: Diagnosis of stones, cysts, and tumours. *arXiv preprint*. arXiv:2502.04367.
[17]Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F. and Adam, H. (2018). Encoder-decoder with atrous separable convolution for semantic image segmentation. In: *Proceedings of the European Conference on Computer Vision (ECCV)*; 801–818.
[18]Liang, G., Fan, W., Luo, H. and Zhu, X. (2022). Improving calibration and reliability of deep neural networks for medical image analysis: A survey. *IEEE Transactions on Artificial Intelligence*.
[19]Mukhoti, J., Kirsch, A., van Amersfoort, J., Torr, P.H.S. and Gal, Y. (2023). Deep deterministic uncertainty for medical image classification. In: *Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)*, pp. 24384–24394.
[20]Hatamizadeh, A., Nath, V., Tang, Y., Yang, D., Roth, H.R. and Xu, D. (2022). UNETR: Transformers for 3D medical image segmentation. In: *Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)*; 574–584.
[21]Kumar, G.M., Satyanarayana, P., Sridhar, B., Rohith, K.T., Kumar, V.V. and Gokula Krishnan, V. (2025). Early Stage Oral Cancer Detection using Machine Learning and Deep Learning Algorithms. In: *Proceedings of the 2025 4th IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics (ICDCECE)*.
[22]Krishnan, V.G., Saleem, P.A.A., Sathyamoorthy, K., Priya, K.H. and Kumar, T.K.S. (2023). Colorectal Cancer Prediction using ResNet-CNN Classification Method. *International Journal of Advances in Soft Computing and its Applications*. 15(2); 49–64.