Deep Learning for Lung and Colon Cancer Detection with Explainable Artificial-Intelligence(AI)
- Version
- Download 18
- File Size 701.17 KB
- File Count 1
- Create Date 31 January 2026
- Last Updated 31 January 2026
Deep Learning for Lung and Colon Cancer Detection with Explainable Artificial-Intelligence(AI)
Jyothsna Priya Yada, Abhigna Thanniru, D.Likhitha, Vasavi College of Engineering,
Hyderabad, Telangana, India
yadajyothsna04@gmail.com, abhignathanniru@gmail.com,likhithadulla9@gmail.com
Abstract: Cancer diagnosis using histopathological imaging is an important task for medical imaging that requires a deep, accurate and interpretable learning model. This study proposes a deep learning framework that employs the Inception Resetv2 model with LBP feature extraction (local binary pattern) to improve detection of lung and colon cancer. Our approach combines folding networks (CNNS) with handmade texture features to improve the classification performance of published histopathological data records. To ensure transparency and confidence in AI control decisions, we integrate explanatory KI (XAI) technologies,including Shapley Additive Description (SHAP) and LOCAL Interpretable Model Aggregation Declaration (LIME), to provide model interpretability. The proposed method achieves high accuracy, accuracy, recall, AUC-ROC results, surpassing traditional models. Characteristic visualiza- tions illustrate key areas that influence predictions and provide insight into deep learning decision making. This study closes the gap between the accuracy and explanation of medical AI and provides a robust and interpretable solution for cancer detection.
Index Terms: Deep Learning, Lung Cancer, Colon Cancer, Histopathology, Explainable AI, Inception-ResNetV2, SHAP, LIME.
Download