This AI can predict a patient’s length of stay at the hospital, time of discharge and even the time of death.
The neural network developed by Google can make predictions regarding the health of a patient by using electronic health record data of the patient. The deep learning based model uses an immense amount of data, such as a patient’s vitals and medical history to make its predictions.
Algorithm to predict death
The new algorithm lines up previous events of each patient’s records into a timeline, which allows the deep learning model to pinpoint future outcomes including time of death. The neural network even includes handwritten notes, comments and scribbles on old charts to make its predictions.
By using these predictions, the hospitals could find new ways to prioritize patient care and adjust treatments plans. Doctors can catch medical emergencies even before they occur. This AI could also free up healthcare workers, who would no longer have to arrange data into a standardized and legible format.
Concerns regarding health data available to single organisation
Currently, health data is uploaded to centralized computer systems but most of these databases exist independently and are spread across various healthcare systems and government agencies. But for this technology to work accurately, all the data must be in one database. Funnelling all of this personal data into a single predictive model owned by one of the largest private corporations in the world is a solution, but it’s not an appealing one.
Some people believe that giving electronic health record of millions of patients in hands of a small number of private companies could give them a chance to exploit health industries and become a monopoly in healthcare. For the implementation of technologies like this, an effective transparent regulatory framework is essential to monitor the activities of private firms.
The author is the Editor-in-Chief of Open Source For You magazine.