AI for Transplant Healthcare

We’re excited to share that our team will be presenting at the World Transplant Conference (WTC) 2025, held in San Francisco, CA, from August 2–6! 

This year, we have five abstracts accepted that highlight our ongoing work using artificial intelligence to support organ transplant decision-making. If you’re attending WTC, we’d love for you to stop by and connect with us during our poster presentations: 

In the first poster session on Sunday, August 3, 2025, at 3:30 PM PST in Exhibit Hall AB, our team has five posters. We will be presenting these in the Kidney section: 

• P1.07.10: “AI-Enabled Approach to Understand the Impact of Deceased Donor Characteristics on Kidney Disposition.”
  Authors: L. Ashiku, R. Threlkeld, T. Levanos, V. Abrol, L. Speir, G. Marklin.
• P1.07.33: “Integrating Surgeon Expertise and AI: The Transplant Surgeon Fuzzy Associative Memory Model for Enhanced Kidney Transplant Decision-Making.”
  Authors: R. Dzieran, C. Dagli, H. Randall, R. Marley, C. Canfield, K. Lentine, M. Schnitzler.
• P1.07.46: “Evaluating Potential Gender Bias in Kidney Discard Prediction Leveraging the Deceased Donor Allocation Model.”
  Authors: A. A. Akinola, O. Fadaiya, R. Threlkeld, C. Dagli.
• P1.07.48: “Assessment of Potential Bias in Kidney Transplant Acceptance Decision Using the Deceased Donor-Recipient Attributes Matching AI Model.”
  Authors: O. Fadaiya, A. A. Akinola, C. H. Dagli, R. Threlkeld.

As well as the Psychosocial & Ethics section: 

• P1.13.48: “Disagreements in AI Adoption for Organ Procurement Organizations: Source of Efficiency or Inequity?” 
  Authors: A. Babaee, D. B. Shank, C. Canfield, J. G. Hall, K. Lentine, H. Randall, H., & M. Schnitzler.  

We’re thrilled to join the global transplant community at WTC to share our findings and learn from others committed to advancing transplant science and practice. If you’re attending, come find us—we’d love to connect, answer questions, and discuss future opportunities for collaboration. 

Stay tuned for updates and reflections from the conference! 

On June 27th, our team participated in the Bioinnovation and Medical Engineering Symposium at the Innovation Lab on Missouri S&T’s campus. The event provided an exciting opportunity to share our research and connect with students, faculty, and professionals interested in the intersection of healthcare and engineering.

The symposium was a great chance to engage with others passionate about biomedical innovation and explore how our AI-focused research can contribute to improving outcomes in organ transplantation. Check out our team below!  

Several members of our team recently had the opportunity to visit Mid-America Transplant, where we presented our progress. Following the meeting, we toured their facilities to gain a deeper understanding of their operational processes and what goes on behind the scenes. This visit provided important insights that will inform our design approach for the upcoming SimUNet trial. We are grateful to MidAmerica Transplant for their time and ongoing collaboration.  

At the National Innovation Forum hosted by The Alliance, Dr. Casey Canfield from Missouri S&T presented our team’s work on designing appropriate AI for organ allocation and acceptance—AI that not only predicts outcomes but also aligns with the real-world needs of transplant stakeholders.

She summarized the three deep learning models that help predict organ utilization, provisional acceptance, and final acceptance. These tools aim to support both Organ Procurement Organizations (OPOs) and transplant centers, particularly in time-sensitive or uncertain decision scenarios.

The core focus of the presentation was that AI must be fair and explainable. The team is actively investigating bias in training data to incorporate stakeholder input to define fairness more inclusively. She also highlighted the importance of explainable AI (XAI) to help users understand and trust the system.

This work is building toward a 2026 field trial using SimUNet, a behavioral science platform that mimics real-world transplant decision-making environments.

To see Dr. Canfields full presentation slides, please click below.

You can watch Dr. Canfield’s presentation—along with the full day of sessions—here!

This research explores the gap between how fairness is defined in AI systems and how it is understood by people who interact with those systems—especially in high-stakes situations like kidney transplant decisions. In machine learning, fairness is often measured using different mathematical formulas that aim to treat groups equally based on factors like race or gender. But for patients, doctors, and families, fairness can feel much more personal and emotional. For example, a system might treat all demographic groups equally on paper but still suggest a lower-priority transplant match for a patient that clinicians feel deserves more urgent care. This disconnect shows that even if an algorithm is “fair” by technical standards, it may not align with what people experience as fair in real life. Our work aims to bridge that divide—combining data-driven fairness with human-centered understanding. To read more about our definition of fairness, check out this blog post here.

To better understand the gap between technical and perceived fairness, we developed an approach that captures how people respond to AI predictions. In our model, non-expert stakeholders are asked whether they agree or disagree with transplant-related decisions made by the system. This simple feedback helps us estimate how fairness is experienced across different social groups. 

We tested this approach through a survey experiment with 85 participants on Prolific. By comparing their responses to standard algorithmic fairness metrics, we uncovered a significant mismatch: even when the model met technical fairness standards, many participants still felt the outcomes were unfair. This reinforces the idea that human perceptions of fairness often go beyond what mathematical definitions can capture.  

To address the disconnect between algorithmic and perceived fairness, we propose two key strategies designed to surface and incorporate stakeholder expertise: 

• Gamification: Interactive experiments that allow stakeholders to engage with transplant decision scenarios, helping researchers understand how people perceive fairness in different contexts. 

• Stakeholder Audits: Sessions where patients, clinicians, and others directly evaluate AI-generated outcomes, offering critical insights into what they see as fair or unfair. 

Rather than aiming to simply educate stakeholders, our goal is to learn from their lived experiences and decision-making values. By centering their perspectives, we aim to build transplant AI systems that align more closely with real-world expectations of fairness that are more inclusive, trustworthy and responsive. Moving forward, we plan to further refine these strategies through continued engagement and implementation in real clinical settings.  

AI adoption in healthcare is rapidly increasing, yet challenges in trust and usability remain. Explainable AI (XAI) seeks to improve transparency, trust, and decision-making by providing users with clear, interpretable insights into AI-generated predictions. A recent stakeholder-driven scoping review, led by Hari Subramanian, explores key design principles for XAI in kidney transplantation, combining stakeholder engagement and literature analysis to develop AI systems that align with real-world medical needs. 

Stakeholder Engagement Approach: 

This study incorporated insights from 39 transplant professionals and patients over a 9-month period through workshops and interviews. Participants included transplant center professionals, Organ Procurement Organizations (OPOs), and kidney recipients, who assessed AI decision-support mock-ups and XAI interfaces to provide feedback. 

Key Findings on XAI Design:  

1. Contextual Use of AI Predictions

• AI can serve as a screening tool, alert system, or second opinion depending on user needs. 
• OPOs prefer AI to be used as a screening tool to optimize kidney placement, while transplant centers prefer to use AI as a decision-support system for a second opinion.

2. Information Needs in AI Predictions 

• AI should provide global (system-level) and local (prediction-specific) explanations. 
• Surgeons prefer system-level binary predictions with confidence scores, while OPOs and recipients favor local explanations that outline the influencing factors.

3. Personalization for Different User Groups 

• Experience level matters: Novices prefer more detailed AI explanations, while experts prefer optional AI explanations.
• Risk tolerance varies: Some surgeons want more AI-generated insights before making decisions, while others prefer minimal AI input. 

4. Customization for Case Complexity 

• AI should offer detailed explanations for complex cases while simplifying insights for routine decisions. 
• Users favored expandable options to access explanations as needed rather than being overloaded with details by default. 

Conclusion: 

This research highlights the importance of stakeholder-driven XAI design in kidney transplantation. AI systems should be flexible, user-controlled, and context-aware, ensuring they enhance—rather than disrupt—clinical workflows. Moving forward, balancing transparency, usability, and trust will be essential in optimizing human-AI collaboration in medical decision-making. Check out the full paper here!

Citation: H Subramanian*, C Canfield, D. B. Shank. 2024. Designing Explainable AI to Improve Human-AI Team Performance: A Medical Stakeholder-Driven Scoping Review. Artificial Intelligence In Medicine. 149: 102780. https://www-sciencedirect-com.mst.idm.oclc.org/science/article/pii/S0933365724000228.

While AI has the potential to enhance productivity in organizations, some still hesitate to adopt it and are cautious about incorporating it into their daily operations. This reluctance is especially pronounced in sectors such as healthcare, where decision-making carries significant consequences. As part of this project, we interviewed OPO Executive Directors and COOs to understand better the process of integrating a new technology into their operations. The expanded technology, organization, and environment (TOE) framework outlines the primary factors that influence the organizational adoption of new information technologies; we used this framework to structure our interview process.

The Technology-Organization-Environment (TOE) framework is a theoretical model used to explain how organizations adopt and implement new technologies by examining three key dimensions: technology, organization, and environment. The technology context considers the characteristics of the technology, such as its relative advantage, complexity, and compatibility with existing systems. The organization context focuses on internal factors like the size of the organization, top management support, available resources, and an innovative culture. The environment context examines external influences, including market trends, regulatory requirements, competitive pressure, and the availability of external support infrastructure. Most of the time, technology adoption follows a bottom-up approach, where employees or departments experiment with and  advocate for new tools or systems before gaining formal approval from leadership. After the technology has proven its validity, companies will share the information with each other, and slowly, more and more companies will adopt the technology until universal adoption. This is a typical innovation curve starting with early adopters. Together, these dimensions provide a comprehensive view of the factors driving technology adoption, making the TOE framework a widely applicable tool across various industries and innovations.

Two contrasting approaches to the adoption process for AI adoption in transplant emerged. One perspective advocates for top-down, national-level coordination of AI system implementation to ensure consistency across the country. However, a bottom-up approach, aligned with the innovation adoption curve and existing diffusion of innovation theories, also surfaced. This analysis indicated that a bottom-up strategy can result in uneven adoption, leading to equity concerns. Variations among OPOs—such as organization size, available resources, and AI literacy—are likely contributing factors to this uneven adoption and implementation. This has implications for how regulators think about encouraging technological innovation.

We are happy to announce the tentative schedule for the AI Ethics Symposium on November 22nd, 2024. We will start the morning with presentations from our three plenary speakers: Alison L. Antes of Washington University School of Medicine, Benjamin Collins of Vanderbilt University Medical Center, and S. Matthew Liao of NYU School of Global Public Health.

Later in the afternoon, we will have parallel sessions for those whose abstracts were approved. Please take a look at the schedule for the afternoon panels below. We can’t wait to see you there.

For more information, check out our page: Symposium Information

If you have not yet registered, please register here: Registration Form

• Registration closes on November 1st, 2024.

Afternoon panel presentation sessions: