Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Online Access: | https://doi.org/10.5281/zenodo.17357263 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866901913453199360 |
|---|---|
| author | Poonam Sahibani, Anilkumar Munani |
| author_facet | Poonam Sahibani, Anilkumar Munani |
| contents | <p>Explainable Artificial Intelligence (XAI) is vital for enabling trust, transparency, and accountability in medical decision support systems. This paper provides a conceptual survey of core XAI techniques— such as LIME, SHAP, Integrated Gradients, Grad-CAM, counterfactual explanations, and concept activation—and analyzes their theoretical foundations, strengths, and limitations in healthcare settings. We further discuss major challenges unique to clinical deployment: explanation validation, human-AI interaction, trust calibration, and regulatory compliance. To guide researchers and practitioners, we propose a conceptual framework mapping XAI methods by clinical task (diagnosis, prognosis, treatment decision), data modality (tabular, imaging, multimodal), and explanation scope (local vs global). We conclude with recommendations for future work: benchmark datasets for explanation evaluation, multimodal interpretation, human‐centered studies, and integration with causal reasoning.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17357263 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Explainable Artificial Intelligence in Healthcare: Methods, Challenges, and a Conceptual Framework Poonam Sahibani, Anilkumar Munani <p>Explainable Artificial Intelligence (XAI) is vital for enabling trust, transparency, and accountability in medical decision support systems. This paper provides a conceptual survey of core XAI techniques— such as LIME, SHAP, Integrated Gradients, Grad-CAM, counterfactual explanations, and concept activation—and analyzes their theoretical foundations, strengths, and limitations in healthcare settings. We further discuss major challenges unique to clinical deployment: explanation validation, human-AI interaction, trust calibration, and regulatory compliance. To guide researchers and practitioners, we propose a conceptual framework mapping XAI methods by clinical task (diagnosis, prognosis, treatment decision), data modality (tabular, imaging, multimodal), and explanation scope (local vs global). We conclude with recommendations for future work: benchmark datasets for explanation evaluation, multimodal interpretation, human‐centered studies, and integration with causal reasoning.</p> |
| title | Explainable Artificial Intelligence in Healthcare: Methods, Challenges, and a Conceptual Framework |
| url | https://doi.org/10.5281/zenodo.17357263 |