Google and the University of California San Francisco recently worked in conjunction to create a novel machine learning tool that can protect against prescription errors that could harm medical patients. The machine learning tool is able to model typical prescription patterns using patient records as data points. The two organizations also stated that the technology works in a similar capacity to fraud detection software that is also powered by machine learning.
Google indicated the tool was accurate three-quarters of the time, and it was 93% accurate when predicting one medication choice for a patient. Overall the tool could protect against instances of prescription error that could harm patients, and the organizations will continue optimizing the system.
Image Credit: Shutterstock