# Comparative evaluation of CAM methods for enhancing explainability in veterinary radiography

## TL;DR

This paper compares different CAM methods to improve explainability in veterinary X-ray image analysis, finding that some methods perform better but none significantly boost diagnostic confidence.

## Contribution

The study provides a systematic evaluation of 11 CAM methods in veterinary radiography, identifying top-performing techniques for visual interpretability.

## Key findings

- EigenGradCAM achieved the highest mean score among CAM methods for veterinary radiography.
- Certain CAM methods provided better visual cues for specific pathologies but did not significantly improve diagnostic confidence.
- FullGrad and XGradCAM had the lowest scores in terms of heatmap relevance and interpretability.

## Abstract

Explainable Artificial Intelligence (XAI) encompasses a broad spectrum of methods that aim to enhance the transparency of deep learning models, with Class Activation Mapping (CAM) methods widely used for visual interpretability. However, systematic evaluations of these methods in veterinary radiography remain scarce. This study presents a comparative analysis of eleven CAM methods, including GradCAM, XGradCAM, ScoreCAM, and EigenCAM, on a dataset of 7362 canine and feline X-ray images. A ResNet18 model was chosen based on the specificity of the dataset and preliminary results where it outperformed other models. Quantitative and qualitative evaluations were performed to determine how well each CAM method produced interpretable heatmaps relevant to clinical decision-making. Among the techniques evaluated, EigenGradCAM achieved the highest mean score and standard deviation (SD) of 2.571 (SD = 1.256), closely followed by EigenCAM at 2.519 (SD = 1.228) and GradCAM++ at 2.512 (SD = 1.277), with methods such as FullGrad and XGradCAM achieving worst scores of 2.000 (SD = 1.300) and 1.858 (SD = 1.198) respectively. Despite variations in saliency visualization, no single method universally improved veterinarians’ diagnostic confidence. While certain CAM methods provide better visual cues for some pathologies, they generally offered limited explainability and didn’t substantially improve veterinarians’ diagnostic confidence.