Spinal disorders are exceedingly common across the population. Over the past 50 years, spine surgery has advanced tremendously and now many degenerative, traumatic and cancerous conditions of the spine can be safely and effectively treated with surgery leading to significant improvements in quality of life and life expectancy for patients with spinal conditions. Despite this, the indications, utilization, techniques, outcomes and economic costs of spine surgery are currently highly variable across the US and the rest of the world, and there remain many areas of inefficiency in providing optimal surgical care to patients. In this setting, artificial intelligence can be thought of as a tool for potentially useful applications that provide timely accurate diagnosis, in addition to standardized and evidence-driven treatments on a large scale at low cost. 

To advance this goal our lab has two major concurrent projects:

  • The first is the investigation of the application of deep learning on radiographic imaging for diagnosis and treatment planning. 
  • The second goal is the study of the use of statistical modelling and machine learning in spine healthcare economics, and in particular whether artificial intelligence can advance personalized, precision medicine in addition to identifying novel and previously not hypothesized predictors of outcome and cost.