Dr. Rachel L. Dzieran, Founder and CEO of AI Amplified Systems, is lead author of a new open-access systematic review in Current Transplantation Reports (Springer Nature) examining how AI has moved from attempting to replace clinical judgment toward genuinely collaborating with it — and what still stands in the way.

The paper, Evolving AI Integration in Complex Medical Decision-Making and Multidisciplinary Transplant Care: A Systematic Review of Human-AI Collaboration, was co-authored with Dr. Cihan H. Dagli and Dr. Robert J. Marley of Missouri University of Science & Technology and published online May 22, 2026.

What the review covers

Following PRISMA 2020 guidelines, the authors screened 2,847 records across eight databases (PubMed/MEDLINE, Embase, Cochrane, IEEE Xplore, Web of Science, Scopus, CINAHL, and PsycINFO) and synthesized 85 studies published between 2019 and 2025. Transplant surgery serves as the exemplar domain: decisions are irreversible, data streams conflict, time pressure is severe, and borderline cases are the norm rather than the exception.

Key findings

  • Collaboration outperforms either party alone. Evidence from randomized controlled trials and multicenter validation studies shows that clinician-AI teams achieve better outcomes than clinicians or algorithms working independently.
  • The accept/reject paradigm is the bottleneck. The strongest results emerge when physicians weigh an AI’s assessment against their own judgment case by case, rather than being forced into a binary choice. Most deployed systems still impose the binary model.
  • Clinicians calibrate trust like they would with a colleague. Physicians read indirect reliability cues — consistency, hesitation, abstention — and adjust accordingly. Systems designed to surface those cues may matter more than exhaustive algorithmic explanation.
  • Binary logic cannot capture surgical reasoning. Transplant decisions live on a spectrum that yes/no classifiers discard. Fuzzy logic systems can formally represent these graded assessments while preserving the interpretability surgeons require to document and defend a decision.
  • Local context is not optional. What works at one transplant center may not transfer to another. The review identifies fuzzy associative memory approaches as a way to encode institution-specific expertise directly into the reasoning process and flag known biases in clinical inputs.

Why it matters to our work

This review formalizes the premise behind AI Amplified Systems: interpretability is a prerequisite for effective human-AI teaming in high-stakes decisions, not a feature added afterward. Our FAME (Fuzzy Associative Memory Engine) platform is built on the same foundation the review points to—interpretable decision support that lets domain experts shape the system’s reasoning rather than simply approve or veto its output.

The paper’s findings apply well beyond transplant medicine. Any complex setting that combines irreversible consequences, ambiguous data, and the need for a human decision-maker to explain the call — clinical, operational, or defense — faces the same design problem.

“AI in healthcare was first imagined as a replacement for clinical judgment. What the evidence actually shows is that the best outcomes come from systems built to work with physicians — and that means building for nuance, not just accuracy.”

— Dr. Rachel L. Dzieran, Founder & CEO

Read the paper

The article is open access.

Citation: Dzieran, R.L., Dagli, C.H., & Marley, R.J. (2026). Evolving AI Integration in Complex Medical Decision-Making and Multidisciplinary Transplant Care: A Systematic Review of Human-AI Collaboration. Current Transplantation Reports, 13(1), Article 20. https://doi.org/10.1007/s40472-026-00519-6

Read the full article on Springer Nature Link:

https://link.springer.com/article/10.1007/s40472-026-00519-6

AI Amplified Systems, LLC is an Ohio-based, veteran-owned company commercializing peer-reviewed research in interpretable AI decision support for defense and health applications.