Michael Kelly, M.D. New Director at Rady Children’s – San Diego

Michael Kelly, M.D. is the new director of Scoliosis and Spinal Deformities at Rady Children’s Hospital-San Diego. Working with Peter Newton, M.D., chief of the Division of Orthopedics and Scoliosis, Dr. Kelly will focus on democratizing medical care via the use of technologies such as artificial intelligence (AI).

Dr. Kelly comes to Rady Children’s from Washington University in St. Louis where he served as associate professor of Orthopedic and Neurological Surgery, director of the Fellowship in Adult and Pediatric Spine Surgery, and director of Spine Research. After earning his M.D. from the University of Massachusetts Medical School, Dr. Kelly did his residency at the University of California – San Francisco. This was followed by two fellowships at the Washington University in St. Louis, where he studied complex adult and pediatric deformities, while simultaneously earning a master’s degree in clinical research.

“I am fortunate to assume this position in a high-functioning spine practice with outstanding partners, Peter Newton and Salil Upasani,” explained Dr. Kelly to OTW. “From a clinical perspective, our first goal is to broaden our regional network and care for more complex and revision pediatric deformity surgeries in California and the West Coast.”

“From an academic perspective, we are going to have weekly meetings for in-depth discussion of cases and literature via formal journal clubs. I have found deep-dive conversations into particular learning cases far more valuable than a review of everything that is going on, where any single case gets only a couple minutes of attention. There is so much to be taught and learned from each other.”

“Similarly, I take an analytical approach to journal club and want trainees to understand study design, methodology, flaws, and THEN conclusions. Too often, people summarize the paper and the conclusions and take no time to review the validity of the approach or interpretation of results.”

“Finally, from a research perspective my primary goal is to begin immunophenotyping our patients, beginning with cerebral palsy, hopefully through a federally funded grant or a foundation grant. The Scoliosis Research Society has funded the early work for us, but we need to work towards clinically deployable methods to risk stratify patients beyond age, body mass index, diagnosis, and other common variables that we have analyzed ad nauseum with little progress.”

OTW asked Dr. Kelly to explain how he plans to incorporate AI into the program. “AI is both simpler and more complex that many people take it to be,” he said.

“AI has a future in our practice as a way to minimize radiation exposure to children. By using machine-learning algorithms we can turn low-dose scans into high resolution images. This will translate to navigation in the operating room and surveillance in the clinic.”

“We hope to work with the Harms Study Group to learn how to use machine-learning, and then AI, to predict optimum 3D reconstructions for scoliosis patients and then bring this to the operating room to achieve our target 95%+ of the time.”

“AI will analyze the immunophenotyping data to bring vast amounts of data which we hope will be clinically relevant and may guide our decision making and care. Finally, we will use machine-learning within the Surgeon Performance Program to help surgeons find their relative strengths and weaknesses so that all surgeons can continue to improve the care they provide.”