New Artificial Intelligence Framework Aims to Increase Heart Transplant Volume by Reducing Unjustified Organ Discards

Dr. Brian Wayda presents AI-driven decision support at ISHLT 2026 to help surgeons utilize donor hearts and reduce the 60% discard rate in the US.

By: AXL Media

Published: Apr 22, 2026, 4:28 AM EDT

Source: Information for this report was sourced from EurekAlert!

Addressing the Critical Shortage Through Data-Driven Allocation

The persistent shortage of heart donors in the United States has left nearly 4,000 patients on waitlists, many of whom remain on life support in intensive care units for months. Despite this dire need, currently only about 30 percent to 40 percent of available donor hearts are successfully utilized for transplantation. According to Dr. Brian Wayda of the NYU Grossman School of Medicine, many hearts are discarded not because they are non-viable, but because the evaluation process is hindered by extreme time pressure and subjective judgment calls. New artificial intelligence tools are being introduced to bridge this gap, offering a consistent and efficient method for transplant teams to make high-stakes, life-and-death decisions.

Accelerating Complex Judgment Under Extreme Time Constraints

When a potential donor heart becomes available, surgical teams typically have a window of only 15 to 30 minutes to review a vast array of medical histories, imaging, and laboratory results. This assessment often occurs in the middle of the night, leading to an environment where clinicians might "anchor" their decision on a single perceived risk factor, such as the donor's age or a history of substance use. AI systems are uniquely suited to support these decisions by providing a rapid synthesis of multiple variables simultaneously. This allows doctors to move past individual "red flags" and view the donor's profile through the lens of objective historical data and national success rates.

The Role of TOPHAT in Benchmarking Donor Risk

A primary innovation in this field is the Tool Predicting Heart Acceptance for Transplant, known as TOPHAT, developed by Dr. Wayda and Dr. Kiran Khush. This web-based model utilizes 20 specific donor characteristics to estimate the probability of an organ's acceptance based on broad clinical data. Rather than labeling a heart as "good" or "bad," the tool provides a comparative analysis against national experience. For example, a heart from an older donor might initially appear high-risk, but the AI can demonstrate that, when considering all other variables, the organ is no riskier than those currently being used in successful transplants.