Science, Safety and Ethics: AI at the Miller School of Medicine

Female physician interacting with a virtual screen
Article Details
  • Dr. Fernando Collado-Mesa has been named the Miller School’s associate vice chair of AI research and ethical use.
  • His work for the Department of Radiology acknowledges integrating AI into radiology requires human oversight and well-defined workflows to enhance patient care while maintaining clinical standards.
  • In his new role, Dr. Collado-Mesa will be charged with ensuring locally deployed AI solutions are free of bias and involve data sets that reflect patient populations.

The rise of artificial intelligence (AI) and machine learning in today’s health care landscape has created a growing need for professionals who understand technology tools and benefits.

The Department of Radiology at the University of Miami Miller School of Medicine is addressing this demand with the appointment of Fernando Collado-Mesa, M.D., as associate vice chair of AI research and ethical use.

“Dr. Collado-Mesa’s contributions over the past few years have been extraordinary and have significantly advanced the university’s position in AI research on both national and international levels,” said Alexander McKinney, M.D., professor and chair of the Department of Radiology at the Miller School and chief of service of radiology at UHealth — University of Miami Health System and Jackson Health System. “His dedication and leadership in this field have propelled our initiatives forward, fostering a culture of innovation and ethical practice within our academic community.”

Dr. Collado-Mesa will continue to serve as a professor of clinical radiology within the department. His research focuses on AI in diagnostic radiology and breast imaging.

“This appointment is a great honor,” Dr. Collado-Mesa said. “It reinforces my commitment to conducting AI research following the highest scientific, safety and ethical standards.”

Guidelines in AI Use

Although this role is new for Dr. Collado-Mesa, he has expertise in researching and implementing AI solutions, both in clinical and educational settings. The AI team in his department follows a phased approach, starting with Institutional Review Board-approved pilot research projects to gauge the effectiveness of AI tools. Once validated, the tools are vetted by the AI Governance Committee and implemented on a broader scale.

Dr. Fernanddo Collado-Mesa
Dr. Fernando Collado-Mesa is working to ensure the Miller School uses AI responsibly and ethically.

This process includes a point-of-care AI deployment workflow that acknowledges integrating AI into radiology requires human oversight and well-defined workflows. The approach ensures that each finding is addressed with urgency and precision, enhancing patient care while maintaining clinical standards.

“Our dedicated teamwork and attention to detail in AI have proven effective,” Dr. Collado-Mesa said. “Thanks to these efforts, the University of Miami Health System is one of a handful of U.S. institutions named an American College of Radiology Recognized Center for Healthcare-AI (ARCH-AI). This achievement underscores our commitment to safety, effectiveness and excellence in AI for health care.”

AI and Radiology

As associate vice chair, Dr. Collado-Mesa will help curate, disseminate and moderate discussions on the latest scientific evidence and technological advancements in AI for radiology. He will also support faculty members with their AI projects and encourage team-science collaborations across departments within the health system and the university.

“In the future, our team aims to develop patient-facing AI that can explain images and reports to both patients and providers,” Dr. Collado-Mesa said. “I will collaborate with our leadership team to implement an AI research platform that integrates seamlessly with our radiology systems, including PACS and report dictation system.”

Dr. Collado-Mesa’s efforts support the vision of department leadership on strategic AI implementation. He has contributed to the implementation of best practices, including evaluating commercial AI solutions for local use and fostering collaborations to develop in-house AI systems while upholding rigorous quality and safety standards. Education in AI is another key priority.

“In 2017, we became one of the first radiology residency programs in the U.S. to incorporate AI education into our curriculum,” Dr. Collado-Mesa said. “Since then, we have expanded AI learning materials for medical students and plan to explore the potential of agentic AI to help students learn from imaging.”

The Importance of Ethical AI Use

While AI holds great promise for health care, it also carries risks, such as bias. Improper ethics can lead to faulty datasets and discrimination against underrepresented populations that reinforce rather than allay racial and demographic disparities.

To help mitigate AI bias on locally implemented AI solutions, Dr. Collado-Mesa led the creation of the Department of Radiology’s first AI Governance Committee. The committee adheres to ethical guidelines and ensures AI systems are trained on data that accurately reflects the local patient population, taking into account race, ethnicity, gender and age. The committee also collaborates with the American College of Radiology to monitor the performance of locally implemented AI tools.

“Radiology has always been at the forefront of health care innovation, and AI is no exception,” Dr. Collado-Mesa said. “Our goal is to develop a multidisciplinary approach that merges image-based AI with EHR data, including demographics, pathology, genetics and laboratory medicine, moving us closer to contextual, patient-centered, precision AI.”


Tags: AI, artificial intelligence, Department of Radiology, Dr. Fernando Collado-Mesa, machine learning, medical ethics, technology