Projects applying artificial intelligence (AI) to healthcare took centre stage at the latest Google I/O, with deep learning and machine learning integral to tools designed to improve the health and quality of life of their users.
“Healthcare is one of the most important fields AI is going to transform,” said Google CEO Sundar Pichai in his keynote address at Google I/O 2018. The event saw the unveiling of a number of Google healthcare products based on artificial intelligence. The website MobiHealthNews picked out the three most important of them.
Improved diagnosis of diabetic retinopathy
Pichai began by talking about the tool Google has developed to detect diabetic retinopathy through the interpretation of retina images. This AI-based innovation enables improved diagnosis by spotting anomalies invisible to the naked eye. It could also be used to predict the risk of cardiovascular events among diabetic patients.
“Last year, we announced our work on diabetic retinopathy, a leading cause of blindness, and we used deep learning to help doctors diagnose it earlier,” said the Google CEO, “and we’ve been running field trials since then at … hospitals in India, and the field trials are going really well. We are bringing expert diagnosis to places where trained doctors are scarce.”
Predictive analyses made possible by deep learning
Pichai added that thanks to the analysis of large quantities of data, AI can predict the chance of readmission 24 to 48 hours earlier than traditional methods. Dr Alvin Rajkomar and Eyal Oren from the Google AI research team offered a closer look at the technology on the company’s blog and in a paper published in Nature Partner Journals: Digital Medicine.
Conducted in collaboration with several US universities of medicine, the study used deep learning models to analyse the medical data of 216,221 patients hospitalised for at least 24 hours. The method enables more accurate predictions of mortality rates, the number of readmissions, the length of stays, and additional diagnoses.
Lookout comes to the aid of visually impaired
Another innovation is the machine learning-based Lookout app for the visually impaired, which is coming to the Android Play Store this year in the US. Lookout uses the smartphone camera to view an individual’s surroundings and warn them with audio cues when objects, text, or people are nearby.
Usable without an internet connection, the app features several modes, among them “Home” and “Work & Play”, that enable it to better recognize its environment. As one of the product managers with Google’s Central Accessibility Team explained, the use of machine learning will enable the device to identify what people are most interested in receiving notifications about.
Contact Allianz Partners
USA, Apr 26, 2018
The US Food & Drug Administration (FDA) has just delivered its first authorisation for AI-based equipment to make an autonomous medical diagnosis without the intervention of a doctor.
Mar 28, 2018
A report published by a company called SITA has shown how artificial intelligence could improve baggage tracking in air transport.