Google using machine learning to predict patient health outcomes

Google says it has seen promising results in its collaboration with US hospitals

Google is working with US hospitals in a bid to predict health outcomes using medical data.

The company's research team, Google Brain, is applying machine learning capabilities that Google has used for its consumer-focused products, including Google Translate and Image Search, to its healthcare products.

The researchers are working with hospitals such as Stanford Medicine, UC San Francisco and The University of Chicago, who have shared patient data with the corporation that, Google claims, has been stripped off of personally identifying data.

Katherine Chou, the head of product at Google Brain told CNBC that the company can "improve predictions for medical events that might happen to you".

"We have validated the data and seen promising results," she said, although the results will only be released after a formal review process.

According to Atul Butte, director of the Institute of Computational Health Sciences at UCSF, Google's in-house machine learning expertise was a key reason for UCSF to take part in the collaboration.

He said the project came to light because UCSF has a range of medical data including admissions reports, medical records, diagnoses and lab results but hasn't yet mined this information to make predictions about patient outcomes.

He emphasised that this wasn't a research project but "more of a scientific collaboration around improving the quality of care for patients".

Earlier this week, Google's London-based DeepMind artificial intelligence subsidiary was criticised, along with the Royal Free London NHS Foundation Trust for the ‘legally flawed' transfer of 1.6 million medical records.

Google had claimed that it planned to use the data in an anonymised form in order to help build an application called ‘Streams', to improve the care of patients with chronic kidney disease.

It had claimed that it needed five years' worth of patients records "to analyse trends and detect historical tests and diagnoses that may affect patient care".

However, Dame Fiona Caldicott, the National Data Guardian of the Department of Health, sent a letter to the hospital's medical director Professor Stephen Powis in which she described the legal basis for the transfer of confidential records as "inappropriate".

Computing's IT Leaders Forum 2017 is coming on 24 May 2017. The theme this year is "Going Digital: Why your most difficult customer is your best friend".

Attendance is free, but strictly limited to IT Leaders. To find out more and to apply for your place, check out the IT Leaders Forum website.