Machine learning (ML) is a rapidly developing branch of computer science that allows computers to learn complex tasks and behaviors. This enables computers to perform many real-world tasks that have hitherto resisted automation. ML is no longer a matter of futuristic fiction or mere laboratory curiosity. Recent advances have made it possible to tackle significant real-world problems, and ML is already making a significant impact in almost every aspect of modern life: from smart phones to high finance, from fuel injection systems in cars to cars that drive themselves, from computers that speak to robots that help build other robots, and from law enforcement to warfare. In today’s world, any smart machine uses some form of ML. Thus, every walk of human life that was once the exclusive domain of extensively trained, highly skilled human professionals is either already being impacted or stands to be impacted by ML. Not surprisingly, ML already has a substantial presence in medicine. In addition to being used for more mundane tasks such as bookkeeping and dispensation, it is also being used to digest vast amounts of cancer research data, to help customize cancer therapies for individual patients, analyze radiological images and other clinical data, and to supervise medical education and training. Applications of ML in neurology are likely to fall under two broad categories: (i) assisting neurological patients with their daily lives (e.g. helping compensate for sensory or motor deficits), or (ii) assisting neurologists in various tasks (e.g. analyzing of neuroimaging data to help make evidence-based decisions customized for each individual patient). Given the unnervingly immense potential of ML, its impact on medicine in general and neurology in particular is likely to increase in the future.
Buy this Article