Comparative evaluation of CAM methods for enhancing explainability in veterinary radiography.
Authors
Affiliations (7)
Affiliations (7)
- AGH University of Krakow, Faculty of Electrical Engineering, Automatics, Computer Science and Biomedical Engineering, Krakow, 30059, Poland. [email protected].
- University of Applied Sciences Western Switzerland (HES-SO), Institute of Informatics, Sierre, 3960, Switzerland. [email protected].
- University of Padua, Department of Animal Medicine, Production and Health, Legnaro, 35020, Padua, Italy.
- University of Applied Sciences Western Switzerland (HES-SO), Institute of Informatics, Sierre, 3960, Switzerland.
- University of Geneva, Medical Faculty, Geneva, Switzerland.
- The Sense Research and Innovation Center, Sion, Lausanne, Switzerland.
- AGH University of Krakow, Faculty of Electrical Engineering, Automatics, Computer Science and Biomedical Engineering, Krakow, 30059, Poland.
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.