Explainable Deep Learning in Medical Imaging: Brain Tumor and Pneumonia Detection

2025-10-21

AI Summary

Medical deep learning models often lack sufficient interpretability, hindering clinical trust. To address this, we propose a Grad-CAM–based interpretability analysis framework to systematically compare the decision-making mechanisms and diagnostic performance of ResNet50 and DenseNet121 on brain tumor (MRI) and pneumonia (X-ray) classification tasks. Results show that DenseNet121 achieves higher accuracy (94.3% for brain tumors; 89.1% for pneumonia) and generates more precise Grad-CAM heatmaps—consistently focusing on pathologically relevant lesion cores—outperforming ResNet50’s broader, less specific attention patterns. Radiologist evaluation confirms the clinical plausibility of these findings. This work is the first to empirically demonstrate, within a unified interpretability framework, that architectural choice critically influences the transparency and clinical credibility of medical AI models. It provides evidence-based guidance for model selection and deployment in trustworthy medical AI systems.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Transparent, Interpretable, Explainable ML
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

Deep Learning (DL) holds enormous potential for improving medical imaging diagnostics, yet the lack of interpretability in most models hampers clinical trust and adoption. This paper presents an explainable deep learning framework for detecting brain tumors in MRI scans and pneumonia in chest X-ray images using two leading Convolutional Neural Networks, ResNet50 and DenseNet121. These models were trained on publicly available Kaggle datasets comprising 7,023 brain MRI images and 5,863 chest X-ray images, achieving high classification performance. DenseNet121 consistently outperformed ResNet50 with 94.3 percent vs. 92.5 percent accuracy for brain tumors and 89.1 percent vs. 84.4 percent accuracy for pneumonia. For better explainability, Gradient-weighted Class Activation Mapping (Grad-CAM) was integrated to create heatmap visualizations superimposed on the test images, indicating the most influential image regions in the decision-making process. Interestingly, while both models produced accurate results, Grad-CAM showed that DenseNet121 consistently focused on core pathological regions, whereas ResNet50 sometimes scattered attention to peripheral or non-pathological areas. Combining deep learning and explainable AI offers a promising path toward reliable, interpretable, and clinically useful diagnostic tools.

Problem

• Developing explainable deep learning for medical image diagnosis
• Improving interpretability of brain tumor and pneumonia detection models
• Addressing clinical trust barriers through transparent AI decision visualization

Innovation

• Used ResNet50 and DenseNet121 for medical image classification
• Integrated Grad-CAM for explainable heatmap visualizations
• Trained models on Kaggle datasets for brain and pneumonia detection