Explainable vision transformer framework for breast lesion classification with Grad-CAM support
8 affiliations
- 1School of Computer Science & Engineering, IILM University, Greater Noida, India
- 2Department of Computer Science, College of Computer Science, Applied College Tanumah, King Khalid University, Abha, Saudi Arabia
- 3Department of Public Health, College of Applied Medical Sciences, King Khalid University, Khamis Mushayt, Almahalah, Abha, 62561 Saudi Arabia
- 4School of Computer Science and Engineering, IILM University, Gurugram, India
- 5School of Computing Science & Engineering, Sharda University, Greater Noida, India
- 6School of Accounting and Finance, Faculty of Business and Law, Taylor’s University, Subang Jaya, Selangor Malaysia
- 7Department of CSE(AIML), JAIN (Deemed to be University), Bangalore, Karnataka India
- 8Faculty of Computer Science, Kardan University, Kabul, Afghanistan
Abstract
Background: Given that breast cancer remains the most lethal malignancy among women globally, there is an urgent need for reliable and comprehensible diagnostic technology. Deep learning models have demonstrated encouraging outcomes in breast imaging; nevertheless, existing methodologies struggle to encompass global context and deliver clinically relevant justifications for their conclusions.
Method: The ExViT-Breast framework was trained and internally validated on CBIS-DDSM (n = 1,566 subjects; masses and calcifications), externally validated on INbreast, and evaluated across modalities on the BUSI ultrasound dataset. A Vision Transformer backbone with a cross-view attention fusion module integrated paired craniocaudal (CC) and mediolateral oblique (MLO) mammographic views. Four explainability methods were compared: Grad-CAM, Grad-CAM++, LayerCAM, and Transformer-LRP. Classification performance was assessed by AUC and sensitivity at one false positive per examination. Explanation quality was evaluated using faithfulness (deletion/insertion curves), clinical plausibility (pointing game against expert-defined lesion regions), and ensemble-based uncertainty estimation. Model calibration was quantified using Expected Calibration Error (ECE).
Results: ExViT-Breast achieved superior diagnostic performance with an AUC of 0.94 ± 0.02 on CBIS-DDSM, significantly outperforming conventional CNN architectures (ResNet-50: AUC = 0.89, p < 0.001) and single-view Vision Transformers (AUC = 0.91, p = 0.003). External validation on INbreast demonstrated robust generalization (AUC = 0.91 ± 0.03), while cross-modality evaluation on BUSI yielded AUC = 0.88 ± 0.04. Transformer-LRP consistently provided the highest explanation quality across all datasets, achieving faithfulness scores of 0.71 ± 0.05 and plausibility rates of 79.5% on CBIS-DDSM, significantly superior to Grad-CAM variants (p < 0.001).
Conclusion: The ExViT-Breast framework demonstrates the potential of explainable Vision Transformers in breast imaging, combining high classification performance with clinically interpretable explanations. The integration of multi-view analysis and quantitative explainability assessment positions this approach as a promising tool for computer-aided diagnosis in breast cancer screening and detection.
Clinical trial: Clinical trial number: not applicable.
Related
- An explainable vision transformer approach for automated neurological disorder classification from brain MRI scans
- An explainable hybrid transformer–CNN framework for multi-class brain tumor classification from MRI images
- A segmentation-guided CNN–Vision transformer feature fusion framework for multi-class breast ultrasound image classification
- ReViT-RF: A hybrid ResNet-50–Vision Transformer with a Radiomic Feature Fusion framework for lung tumor classification and TNM-based stage prediction using CT images
- MedFuse: dual-stream fusion of convolutional and vision transformer-based features for enhanced medical image classification
BibTeX
@article{kumar2026explainable,
title = {{Explainable vision transformer framework for breast lesion classification with Grad-CAM support}},
author = {Ajay Kumar and Mudassir Khan and Gunjan Mittal and Izhar Husain and Sambhavi Shukla and Syed Arshad Ali and Anu Sayal and Janhvi Jha and Riaz Ahmad Ziar},
journal = {Cancer Imaging},
year = {2026},
month = jul,
doi = {10.1186/s40644-026-01097-7},
url = {https://doi.org/10.1186/s40644-026-01097-7},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Explainable vision transformer framework for breast lesion classification with Grad-CAM support
AU - Ajay Kumar
AU - Mudassir Khan
AU - Gunjan Mittal
AU - Izhar Husain
AU - Sambhavi Shukla
AU - Syed Arshad Ali
AU - Anu Sayal
AU - Janhvi Jha
AU - Riaz Ahmad Ziar
PY - 2026
DA - 2026/07/30
JO - Cancer Imaging
DO - 10.1186/s40644-026-01097-7
UR - https://doi.org/10.1186/s40644-026-01097-7
AB - Background: Given that breast cancer remains the most lethal malignancy among women globally, there is an urgent need for reliable and comprehensible diagnostic technology. Deep learning models have demonstrated encouraging outcomes in breast imaging; nevertheless, existing methodologies struggle to encompass global context and deliver clinically relevant justifications for their conclusions. Method: The ExViT-Breast framework was trained and internally validated on CBIS-DDSM (n = 1,566 subjects; masses and calcifications), externally validated on INbreast, and evaluated across modalities on the BUSI ultrasound dataset. A Vision Transformer backbone with a cross-view attention fusion module integrated paired craniocaudal (CC) and mediolateral oblique (MLO) mammographic views. Four explainability methods were compared: Grad-CAM, Grad-CAM++, LayerCAM, and Transformer-LRP. Classification performance was assessed by AUC and sensitivity at one false positive per examination. Explanation quality was evaluated using faithfulness (deletion/insertion curves), clinical plausibility (pointing game against expert-defined lesion regions), and ensemble-based uncertainty estimation. Model calibration was quantified using Expected Calibration Error (ECE). Results: ExViT-Breast achieved superior diagnostic performance with an AUC of 0.94 ± 0.02 on CBIS-DDSM, significantly outperforming conventional CNN architectures (ResNet-50: AUC = 0.89, p < 0.001) and single-view Vision Transformers (AUC = 0.91, p = 0.003). External validation on INbreast demonstrated robust generalization (AUC = 0.91 ± 0.03), while cross-modality evaluation on BUSI yielded AUC = 0.88 ± 0.04. Transformer-LRP consistently provided the highest explanation quality across all datasets, achieving faithfulness scores of 0.71 ± 0.05 and plausibility rates of 79.5% on CBIS-DDSM, significantly superior to Grad-CAM variants (p < 0.001). Conclusion: The ExViT-Breast framework demonstrates the potential of explainable Vision Transformers in breast imaging, combining high classification performance with clinically interpretable explanations. The integration of multi-view analysis and quantitative explainability assessment positions this approach as a promising tool for computer-aided diagnosis in breast cancer screening and detection. Clinical trial: Clinical trial number: not applicable.
SN - 1740-5025
SN - 1470-7330
ER -