A co-authored systematic review of 76 studies examining explainability methods, clinical trust, evaluation, and barriers to deployment.
The research
Published in Algorithms in 2026, this systematic review was written with Sayed Mehedi Azim and Iman Dehzangi. It examines 76 studies published between 2020 and 2025 on explainable AI (XAI) in healthcare.
What the review found
Different data types tend to call for different explanation methods. The review also highlights uneven evaluation of whether explanations are faithful and useful in clinical settings, limited real-world validation, and gaps between technical explanations and regulatory needs.
Why it matters for adoption
A model can be accurate yet difficult to scrutinize or trust. Responsible adoption depends on the people making consequential decisions being able to understand when and how an AI output should be used.
Publication
Ojobo Agbo Eje, Sayed Mehedi Azim, and Iman Dehzangi. Explainable AI Applications in Healthcare: A Systematic Review. Algorithms 19(6), 488 (2026). Read the open-access paper.