JMIR Publications

ABSTRACT

Background:

Artificial intelligence is increasingly being integrated into clinical diagnostics, yet its lack of transparency hinders trust and adoption among healthcare professionals. Explainable AI (XA1) offers a solution by making AI-driven decisions more interpretable and reliable for clinical use.

Objective:

This study seeks to evaluate radiology reports employing Explainable AI (XAI) as a means for improving healthcare practitioners’ confidence, openness, and comprehension of AI-assisted diagnostics.

Methods:

This study employed the Indiana University chest X-ray Dataset containing 3955 textual reports and 7,470 images. Reports being classified as either normal or abnormal by using machine learning algorithms and ensemble methods, deep learning models, transformers and language models (GPT-2, T5, LLaMA-2, LLaMA-3.1). For image processing, models like CNN, DenseNet121, and DenseNet169 were used. The best models were assessed using XAI methods- SHAP and LIME to support medical experts for decision making.

Results:

LLaMA-3.1 model achieved highest accuracy of 98% in classifying the textual radiology reports. Statistical analysis confirmed the model robustness, with Cohen’s kappa (k=0.981) indicating near perfect agreement beyond chance, both chi-square and Fisher’s Exact test revealing a high significant association between actual and predicted labels(p<0.0001). Although McNemar's test yielded a non-significant result (p=0.25) suggests balance class performance. While the highest accuracy of 84% was achieved in the analysis of imaging data using the DenseNet169 model. Explainable Artificial Intelligence methods SHAP and LIME were applied on best models’ predication that provide useful information how the decisions made provided assurance regarding the interpretability of AI-based diagnostic solutions.

Conclusions:

The research underscores that explainability is an essential component of any AI systems used in diagnostics and helpful in the design and implementation of AI in the healthcare sector. Such approach improves the accuracy of the diagnosis and builds confidence in health workers, who in the future will use explainable AI in clinical settings, particularly in the application of AI explainability for medical purposes.