From creating “digital twins” to neuromonitoring improvements, spine surgeons discuss ways artificial intelligence will integrate into their work next.
Editor’s note: Responses were lightly edited for clarity and length.
Question: What will be the next major way AI is integrated into spine care?
Brian Gantwerker, MD. The Craniospinal Center of Los Angeles: AI will undoubtedly have an effect on patient care in spine in the years to come in terms of treatment algorithms. Someone, somewhere will sell a large dataset to a company — if it hasn’t happened already — that will essentially try to tell surgeons how to take care of patients. Whether that is for the good of patients or not is up for debate. The motto of “move fast and break things” is cute and all, but not when peoples’ lives and mobility are at stake. After all, if you are learning to do the trapeze, do you start with no net?
Arthur Jenkins, MD. Jenkins Neurospine (New York City): The next innovation in healthcare is applying artificial intelligence to the neuromonitoring that goes on during an operation by improving the sensitivity and specificity of the SSEPs, EMGs, and MEPs, we will not only get a better selectivity for when there is a neurologic problem versus when it is simply artifactual activity that’s not related to a surgical problem, but also the ability to determine when enough decompression has been performed and that further decompression would be fruitless, i.e. to determine the minimum amount of decompression to achieve equivalent outcomes. This is relevant because many cases have a point at which it’s unclear if you’ve done enough, or past the point of marginal futility. Being able to accurately and adequately say the procedure is completed may effectively shorten the surgical and anesthesia time, which reduces multiple other medical risks, including wound breakdown, pulmonary problems and venous problems, all of which correlate with longer OR times. The opposite end of the spectrum, inadequate decompression, frequently results in residual or recurrent symptoms early in the recovery period, dissatisfaction and potential lawsuits.
Samuel A. Joseph Jr., MD. Joseph Spine Institute, Tampa Bay, FL. I believe AI-assisted surgical planning and navigation is one of the most powerful tools we’ve ever had at our disposal.
The integration of AI with next-generation robotics and advanced imaging allows us to preoperatively model each patient’s unique spinal anatomy in extraordinary detail. Using data from MRI and CT imaging, our AI systems map out optimal trajectories for screw placement, evaluate spinal alignment, and simulate corrective procedures with unparalleled precision.
During surgery, robotic arms guided by AI enhance our steadiness and accuracy, while real-time imaging and augmented reality overlays keep us constantly aligned with our surgical goals. This approach dramatically reduces operating time, lowers complication rates and personalizes care to an extent that was unimaginable just a decade ago.
What makes this technology revolutionary is its ability to turn data into insight. We are no longer working in the dark; we are guided by predictive models that optimize outcomes, preserve neural integrity, and provide patients with a clearer path to recovery.
As we implement these systems at Joseph Spine Institute, we are not just performing surgery — we are creating the future of spine care.
Choll Kim, MD, PhD. Excel Spine (San Diego): AI will be a part of our everyday lives, including our work place where we care for patients with spinal disorders. I hope, and expect, that in the near future, an AI version of myself will be able to attend to the boundless questions our patients, and their families, have about their spinal ailments and their treatment options. I see a Choll Kim AI who is available 24/7, 365 days a year, who is infinitely patient and charming, with a near infallible memory who can attend to a myriad of questions and requests that seem “basic and simple” to us, but for the individual patient can be a source of great confusion and anxiety. If I could address this portion of my patient’s needs, it would greatly free the real me to pay attention to the things that only the real Choll Kim can do. The only fear I have of this future is that many may prefer the AI Choll Kim over the real Choll Kim.
Philip Louie, MD. Virginia Mason Franciscan Health (Seattle): I want to highlight three more unique possible integrations that are not often discussed.
1. Personalized risk profiling beyond clinical data: AI will evolve beyond traditional risk calculators by incorporating psychosocial and behavioral data to provide personalized treatment recommendations. Inspired by recommendation engines in consumer tech, these models could align surgical decisions with individual patient values, expectations, and risk tolerance. Combining perceptions related to predicted outcomes and perceived quality of life.
2. “Administrative” intelligence for workflow optimization: AI has the potential to dramatically streamline the administrative burden in spine care, specifically with automating tasks such as surgical pre-authorization, documentation, and real-time billing code generation. This integration could reduce the rapidly rising administrative burden that we are all facing
3. Digital patient twin for engagement and recovery: Using AI-generated “patient twins” that simulate an individual’s recovery journey based on their unique data, spine care teams can offer more proactive and predictive perioperative engagement along the entire episode of care. These digital companions could deliver tailored education, anticipate complications, and help guide rehab. Maybe this can help close the expectation-outcomes mismatch and drive satisfaction following various treatments (especially surgery).
Roy Vingan, MD. New Jersey Brain and Spine (Paramus): From my perspective, artificial intelligence may impact the spine field in several ways. Regarding imaging interpretation, some radiographic interpretations may become slightly less subjective and, over time, may correlate gradually more with clinical outcomes as massive data can be analyzed more efficiently. Secondly, AI modeling may be more efficiently applied to spinal curvature measurements and assessments, particularly in the context of surgical planning. Doing so may help to improve segmental correction and potentially reduce adjacent segment degeneration, especially in fusion procedures. Eventually, this will be further advanced by potential 3D printing technologies in the operating room specifically designed and individualized for the patient.
Christian Zimmerman, MD. St. Alphonsus Medical Group and SAHS Neuroscience Institute (Boise, Idaho): For the better part of a year, our parent organization and Neuroscience’s Institute have utilized the Digital Analysis Expressions (DAX) program as an efficiency enhancement, communicative tool in our clinics. From a process standpoint, patient-consent is primarily obtained (using a Haiku based application) which then allows real time transcription of both history/physical, data/modality collation and surgical planning to be recorded, transcribed and available contemporaneously. Editing and immediate submission routinely save hours of post-clinic reiteration and time. It has been well received by patients, but more importantly, it has become extremely useful for both staff and advanced providers.
AI enables better data-driven decisions within our health care system. In a digitalized health care environment, the quality of decision-making relies on the availability and accuracy of underlying data. AI can assist in decision-making by offering real-time recommendations based on clinical guidelines or advancements, reducing the likelihood of medical mistakes. For example, IBM Watson Health uses built-in integrations to provide clinical decision support and achieve a high level of agreement with physician recommendations.
