Neurosurgeon Michael Fehlings specializes in complex spine surgery and is the Vice Chair Research for the Department of Surgery at the University of Toronto and Co-Director of the University of Toronto Spine Program. We caught up with him to learn more about where the use of Artificial Intelligence (AI) in neurosurgery could be heading in the future.
What inspired you to specialize in neuroscience and spinal cord injuries?
I first became intrigued with serious disorders of the brain because my grandfather sustained multiple strokes. I saw the impact that brain disease had on him and it robbed him of his language. When I went to medical school I was fascinated with brain disorders and I felt that neurosurgery provided practical solutions to help people like my grandfather.
What's the biggest challenge you face in your work?
One theme that has driven my professional career has been the challenge of trying to optimize repair and regeneration of the injured brain and spinal cord. When I entered medical school, the dogma was that neurons in the central nervous system do not regenerate. Over the last several decades that dogma has been refuted and the possibility of being able to block factors that inhibit regeneration has been identified. The challenge now is to translate these discoveries from molecule to patient.
How did you become interested in AI?
I’ve always been interested in mathematics. For me, statistics as an undergraduate student and later as a graduate student was a natural fit. I always enjoyed the opportunity to manipulate large data sets and discern patterns and correlations with the goal of gaining novel insights. Machine learning and artificial intelligence represent an extension of conventional statistics with the unique appeal that these technologies allow for the examination of large data sets. The databases that are now being generated at the basic and clinical science levels are so large that conventional statistical approaches are lacking. The opportunity now is to apply these powerful mathematical algorithms to solve fundamental questions in medicine and biology.
What excites you the most about the possibilities of AI in neurosurgery?
For me, the next big challenge is about realizing the opportunity for precision-based medicine to tailor treatments and diagnoses to specific patients. The big opportunity here will be to merge precise clinical phenotyping with detailed imaging, including microstructural magnetic resonance imaging, and with genetic approaches, including whole-genome sequencing. Applying AI to these large datasets will enable us to be much more precise with the correct diagnosis and treatment for each individual patient.
What’s the number one piece of advice you’d give to medical students following in your footsteps?
Follow your dreams and don’t let anyone tell you it is impossible. It is possible to combine a busy clinical career in neurosurgery with an active research and academic program, while enjoying family life, children and friendships outside of medicine. Don’t let anyone tell you that it is not.
What exciting projects or initiatives are you currently working on?
We are actively engaged in regenerative neuroscience in my research program and have developed unique bioengineered neural stem cells which are genetically modified to express factors to enhance repair and regeneration while degrading the glial scar. We are using AI and machine learning approaches to better understand the interaction between these neural stem cells and the host central nervous system environment using single-cell RNA-sequencing and sophisticated bioinformatics platforms.
