FURI | Spring 2026

Improving Trust in AI-Based Pneumonia Diagnosis through Explainable Deep Learning

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

This project investigates how explainable deep learning can improve transparency and trust in automated pneumonia detection from chest X-ray images. Convolutional neural networks will be trained to classify medical images, while explainable artificial intelligence techniques, including Gradient-weighted Class Activation Mapping (Grad-CAM) and SHapley Additive exPlanations (SHAP), will highlight regions influencing model predictions. This approach aims to support more accurate and interpretable diagnostic tools for healthcare professionals. Future work will explore expanding the model to additional diseases and improving robustness across diverse medical datasets.