The research scientist from the University of Texas Medical School discussed some of the challenges and opportunities that machine learning presents in patient care.
“One of the biggest challenges that is also an opportunity is to have large data repositories with common data elements, and then have an open-source data distribution model that would allow for computer scientists to use data to solve some really important problems.”
Findings from a recently published study on prognostic machine learning (ML) models suggest that these models may be used to predict delayed cerebral ischemia (DCI) as well as functional outcomes in subarachnoid hemorrhage (SAH) care. In the study, the ML models significantly outperformed standard models (SMs).
Conducted by Jude Savarraj, PhD, research scientist, department of neurosurgery, University of Texas Medical School at Houston, and colleagues, the results show that the ML model outperformed the SM in area under the curve (AUC) by 0.20 (95% CI; -0.02 to 0.4) for DCI, 0.07 (standard deviation [SD], 0.03; 95% CI, 0.0018–0.14) for discharge outcomes. Additionally, the ML model bested the SM model by 0.14 (95% CI, 0.03–0.24) for 3-month outcomes, and matched physician performance in predicting 3-month outcomes.
Initially only using routine variables included in a patients’ electronic medical record (EMR), Savarraj and colleagues noticed that the ML models performed even better when also including clinician-determined Hunt-Hess Scale score in addition to standard EMR variables (AUC, 0.85; SD, 0.05; 95% CI, 0.75–0.92). Only using EMR data yielded an AUC of 0.81 (SD, 0.05; 95% CI, 0.71–0.89; P <.05).
In this interview, NeurologyLive spoke with Savarraj to learn more about the use of machine learning in healthcare and the opportunities for different healthcare systems to pool data for machine learning models.