Google has been working on a new algorithm that will help in the prediction of a patients odds of living.
According to Bloomberg, A woman who was suffering from a late stage of breast cancer saw two doctors and got a radiology scan. The hospital result of this assessment was that the patient had a 9.3% chance of dying during her stay. Google’s new algorithm assessed her records and predicted the death risk to be 19.9 per cent. The patient in question passed away in a few days.
This account was published by Google in may and highlights the ‘health-care potential of neural networks’, which is a kind of artificial intelligence that works with data. This data algorithm improves and learns on its own.
Also, read: Twitter Co-founder Invests in Indian Artificial Intelligence Powered Health Chatbot
This tool that has been created by Google sifts through mounds of existing data that was not earlier available like PDF’s, scribbled notes. The neural net takes in all the information and gives a result that is much faster than existing ways.
It also shows why did the neural net reach that particular result and how.
Doctors, Hospitals and health-care providers have been looking for ways to improve the existing healthcare system, but current ways of mining health data are very expensive.
Google plans to move this system to clinics and in order to predict the health of a patient on the basis of information that is shared.
Since acquiring data in the medical field is hard it is yet to be seen as to when the algorithm/ medical brain will be ready for commercial purposes. However, according to a report at Google, I/O Lily Peng who is a member of Medical Brain indicated that the team’s research outmatched humans in assessing the risk of a heart disease.