• International Open Access Publishing Platform
  • Peer Reviewed Journals
Open Access Review

Explainable AI for Clinical Decision Support: A Systematic Review

DOI: https://doi.org/10.5282/jai.2026.002
Ifeoma Balogun

Published in Journal of Artificial Intelligence · April 9, 2026

Cite Article

Abstract

A systematic review of explainability techniques applied to AI-assisted clinical decision-making, with a focus on trust and adoption.

Keywords

Explainable AI Healthcare Clinical Decision Support

Full Text

Clinical adoption of AI decision-support tools continues to be constrained by clinician distrust of opaque model outputs. This systematic review synthesises 47 studies published between 2020 and 2026 evaluating explainability techniques in clinical AI deployments.

SHAP and LIME-based methods dominate the literature, but the review finds limited evidence that explanation quality alone drives clinician trust; workflow integration and calibrated confidence reporting appear to matter at least as much. The review concludes with a proposed evaluation framework combining technical explainability metrics with clinician-reported trust measures.

References

  1. Lundberg S., Lee S. A Unified Approach to Interpreting Model Predictions (SHAP). NeurIPS, 2017.
  2. WHO. Ethics and Governance of Artificial Intelligence for Health, 2021.