🧬 Breast Cancer Classification Using Pre-trained CNNs with Explainable AI for Enhanced Decision Support

---

🎓 Accepted at:
📌 2025 International Conference on Electrical, Computer and Communication Engineering (ECCE)
🔗 IEEE Xplore
📄 ResearchGate

---

📌 Abstract

Early detection of breast cancer is crucial for effective treatment. This study proposes a deep learning approach using pre-trained CNNs to classify breast cancer and improve decision transparency through Explainable AI (XAI). We utilized an imbalanced BUSI dataset, applied augmentation and preprocessing, and evaluated four advanced CNNs: EfficientNetV2S, InceptionResNetV2, EfficientNetV2M, and XceptionNet.

To enhance model interpretability, we used Faster ScoreCAM and LIME. Among all models, EfficientNetV2S achieved the highest accuracy of 91.02%, demonstrating the potential of explainable deep learning models in medical imaging.

---

🚀 Highlights

• 🩺 Medical Domain: Breast Cancer Ultrasound Images (BUSI Dataset)
• 📈 Model Accuracy: EfficientNetV2S - 91.02%
• 🤖 Deep Learning Models: EfficientNetV2S, InceptionResNetV2, EfficientNetV2M, XceptionNet
• 🧠 Explainable AI: LIME, Faster ScoreCAM
• ⚙️ Balanced Training via Augmentation
• 📊 Transparency in Clinical Decision Support

---

🧠 CNN Architectures Used

Model	Accuracy
EfficientNetV2S	91.02% ✅
InceptionResNetV2	89.65%
EfficientNetV2M	88.74%
XceptionNet	87.93%

---

🧪 Explainability (XAI)

We leveraged LIME and Faster ScoreCAM to visualize which regions of the ultrasound image influenced the model’s prediction:

• ✅ Improves transparency and clinical trust.
• 🎯 Highlights critical regions contributing to the classification.

🌐 Transparency = Trust.

🔹 EfficientNetV2M + Faster ScoreCAM

---

🔹 EfficientNetV2S + LIME

🖼️ Sample Output

---

@INPROCEEDINGS{11012958,
  author={Kabir, Md. Zobayer Ibna},
  booktitle={2025 International Conference on Electrical, Computer and Communication Engineering (ECCE)}, 
  title={Breast Cancer Classification Using Pre-trained CNNs with Explainable AI for Enhanced Decision Support}, 
  year={2025},
  volume={},
  number={},
  pages={1-6},
  keywords={Deep learning;Training;Accuracy;Explainable AI;Source coding;Prevention and mitigation;Decision making;Breast cancer;Data models;Feeds;Medical Imaging;Breast Cancer Classification;CNN;BUSI Dataset;Explainable Artificial Intelligence (XAI);Faster ScoreCAM;LIME;Explainability;Deep Learning},
  doi={10.1109/ECCE64574.2025.11012958}}