Abstract
The clinical neurosciences have historically been at the forefront of innovation, often incorporating the newest research methods into practice. This chapter will explore the adoption, implementation, and refinement of big data and predictive modeling using machine learning within neurosurgery. Initial development of national databases arose from surgeons aiming to improve outcome predictions for patients with traumatic brain injury in the 1960s. In the following decades, other surgical specialties began building databases that left a lasting impact on the current national neurosurgical databases, particularly in spine surgery. Significant contributions to the literature have been made as a result of the numerous registries today, leading to broad quality improvements for neurosurgical patients. Important limitations of large databases do exist, including lack of standardized reporting and challenges in data extraction from medical records. New vistas will include the use of metadata to track human function, performance, and pain in a real-time manner to augment the reliance on traditional patient-reported outcome measures (PROMs). Overall, big data has demonstrated significant utility within neurosurgical research and machine learning-powered analyses have highlighted several promising areas of interest for future exploration.