Role of Artificial Intelligence in Improving the Process of Organ Procurement From Detection to Donation
Organ procurement from patients after brain death faces 2 main challenges: identification of possible donors and interaction with the brain dead donor’s family. During this period, the family sees their loved one in a coma-like condition, supported by advanced devices in the intensive care unit and still hoping for recovery, while the life of their loved one has been completely destroyed and only short-term organic survival continues with these devices. After accepting this difficult reality, the family must make a serious decision to donate the patient’s organs to unknown patients on wait lists. At this stage, the analytical and conversational capabilities of artificial intelligence
can be used as an auxiliary tool alongside organ procurement teams to provide innovative solutions in management of interviews. In our organ procurement unit, we use artificial intelligence as an analytical consultant in case of family refusal. Thus, it is necessary to design and train AI tools based on specialized prompts and based on different consent models (opt-in or opt-out models), as well as that use language in accordance with the laws and cultural requirements of each region. If 2-way interactions continue with such tools, along with recording practical experiences and providing ongoing feedback from organ procurement teams, this technology can continuously learn and improve.
Key words : AI tools, Brain death, Family consent, Possible donor
Dear Editor:
The integration of artificial intelligence (AI) into health care is no longer a future vision but a present reality.1 In organ donation and transplantation, AI has the potential to address persistent inefficiencies faced by organ procurement units (OPUs), including timely donor identification, meaningful 2
Recent studies have described several applica-tions of AI in this field. These include organ allocation, prediction of the likelihood of family consent or refusal, and support of immunosuppressive drug prescriptions.3 Other research has highlighted how AI can improve surgical planning and reduce intra-operative risks, enhance posttransplant outcomes through personalized treatment, and optimize logistics in organ procurement, making the process more efficient.4 More specific models have been developed to predict cases in which families consent but no organs are ultimately recovered to save resources.5 A state-mandated AI system in India also showed value in improving transparency and documentation of brain-death donation pathways, although the system unexpectedly prolonged retrieval times, underlining the importance of workflow integration and iterative refinement.6 In addition, a “responsible AI framework” has been proposed to address systemic barriers by integrating explainability and ethical safeguards.7
Despite these contributions, one area remains underexplored, which is the use of generative AI to support donor coordinators. In our center, we have begun using generative AI to provide tailored information during family approaches. Our expe-rience has suggested that each OPU should develop its own AI tool trained on the cultural, ethical, and demographic characteristics of its local population. We believe this is a novel and important direction that deserves further research.
Beyond family interaction, AI can also assist in donor identification. In our approach, complex generative AI applications do not involve hospital and data analyses or internet infrastructure findings. Instead, nurses in any hospital ward, regardless of their level of experience or specific knowledge about organ donation, enter information of patients with reduced levels of consciousness into the trained AI tool, which evaluates suitability for donation without the influence of personal judgment. This method helps prevent the unnecessary rejection of acceptable cases, such as treated bacterial meningitis, lupus, or other issues, and promotes greater transparency in the donation process.
Family consent remains one of the most sensitive challenges, with rates shown to range between 60% and 85% in our previous work.8 Artificial intelligence systems trained on ethical, religious, legal, and medical principles could assist coordinators during interviews by adapting communication strategies. More experimental approaches are emerging, such as analyses of the donor’s digital footprint to infer values, generation of voice messages from the deceased’s person, or creation of virtual reality simulations of the transplant recipient’s life. Although such tools may help foster empathy and reduce uncertainty, their use must remain voluntary and subject to strict ethical and legal oversight. In this regard, key concerns include postmortem data consent, cultural sensitivity in voice replication, and respect for family autonomy. However, in a simple approach through attention to the diversity of religious rulings across Islamic schools of thought and other faith traditions, training AI tools with real-time information and linking them to authoritative resources from various religious bodies (Muslim, Catholic, or others) can provide coordinators with accurate and timely guidance. This guidance can then be conveyed to families, thereby reducing uncertainty and saving valuable time in the decision-making process.
In summary, AI has already shown promise in organ allocation, surgical planning, logistics, and prediction of family consent. The role of generative AI in directly supporting coordinators and addressing barriers to consent is still a new frontier that requires careful exploration. Regulatory frameworks must evolve to ensure transparency, ethical integrity, and trust as these technologies move from pilot use to broader clinical practice.
Conclusions
Integration of AI offers a paradigm shift in donor identification, family engagement, and organ allocation.
References:

Volume : 23
Issue : 12
Pages : 845 - 846
DOI : 10.6002/ect.2025.0216
From the 1Student Research Committee, Alborz University of Medical Sciences, Karaj, Iran; the 2School of Medicine, Tehran University of Medical Sciences, Tehran, Iran; and the 3Lung Transplantation Research Center, National Research Institute of Tuberculosis and Lung Diseases, Shahid Beheshti University of Medical Sciences, Tehran, Iran
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Fariba Ghorbani, Niavarran, Daraabaad, Masih Daneshvari Hospital, Tehran, Iran
E-mail: dr.f.ghorbani@gmail.com