Artificial intelligence and machine learning have made inroads into the healthcare system, both for administrative processes and clinical applications. Surgeons are also excited about the future potential of AI in spine.
At the American Association of Neurological Surgeons Annual Scientific Meeting in April, researchers discussed how AI could improve referrals for degenerative lumbar spine conditions. The researchers created an AI platform to triage spine patients, based on knowledge from patients who were previously treated for lumbar spine pain from 2013 to 2018. The solution developed an artificial neural network with the data points and identified 55 factors associated with patients who underwent surgery to predict the likelihood of spine surgery progression.
Sessions at the American Academy of Orthopaedic Surgeons Annual Meeting also included artificial intelligence applications, including a session moderated by Joseph Schwab, MD, chief of orthopedic spine surgery at Massachusetts General Hospital in Boston.
"In spine fracture management, a main focus for us is predicting which patients will fracture or not," he said in an interview with Becker's Spine Review. "One of the biggest trends in the field right now is using machine learning or artificial intelligence to predict fractures. This trend will allow for prediction models that are currently relatively static to be replaced with AI to offer a more personalized and accurate description of where fractures will happen on a patient."
AI has the potential to improve preoperative planning, surgical execution and patient-reported outcomes. Spine care providers are already collecting and organizing data with AI systems to predict which interventions are best for different groups of patients.
"The goal is to use data to draw conclusions that contribute to evidence-based medicine and transform care, helping to ensure greater value and better outcomes with reduced overall cost," said Azadeh Farin, MD, a neurosurgeon and spine surgeon at Long Beach (Calif.) Medical Center. "The trend is to use technology to influence physician decision-making, improve clinical outcomes and stabilize institutional finances."
On the orthopedic side, Winston-Salem, N.C.-based Novant Health Orthopedics & Sports Medicine has adopted AI platforms for value-driven care. The health system is partnering with organizations such as Microsoft and Google to build reliable, optimized care pathways.
"As we move more toward this model of care, we not only need robust data collection systems, but tools that can interpret this data effectively," said Bryan Edwards, MD, senior vice president of Novant Health Orthopedics & Sports Medicine. "That's where AI steps in. Our CEO, Carl Armato, and CMO, Eric Eskioglu, MD, have a vision to become leaders in AI by adopting its use into our daily clinical work."
In the future, AI could combine with robotic technology for more accurate spinal instrumentation placement and achieving sagittal balance.
"The addition of artificial intelligence technology would advance the current robotic/navigation technology to adapt easily to intraoperative changes in the spinal anatomy that differ from preoperative images from X-rays, CT scans and MRIs," said Eugene Koh, MD, PhD, associate professor in the department of orthopedics at Baltimore-based University of Maryland School of Medicine and chief of orthopedic spine at Baltimore VA Medical Center. "The current systems are sensitive to any changes in the calibration markers as well as modifications to the spinal anatomy, like interbody devices. Artificial intelligence would allow for the system to adapt with minimal recalibration."