Artificial intelligence in neurosurgery

Main Article Content

Anton Fomenko
Andres Lozano

Keywords

Abstract

Recently, artificial intelligence (AI) has experienced a renaissance of sorts, with applications from autonomous vehicles on our roads to digital personal assistants in our homes. But in fleshing out “AI,” it is the second letter of the term that prompts lively debate, for what exactly defines intelligence? Setting aside philosophical ruminations for the moment, AI can be practically thought of as any technology which simulates the cognitive modules of the biological brain, namely: information gathering, processing, learning, and reasoning.  In medicine, the exponential growth of peer-reviewed literature and complex datasets in the last half-century has begun to saturate the physician’s ability to stay accurately up-to-date. These increased demands on the modern clinician can exacerbate cognitive biases, which are estimated to contribute to 40,500 patient deaths per year from medical errors.1 Through the development of reliable, efficient, bias-free AI systems to assist the surgeon, these unacceptable statistics can potentially be reduced.