Google Health diabetic retinopathy diagnosis

Google Health's diabetic retinopathy diagnosis system uses artificial intelligence to analyse retinal images, with the aim of providing accurate and early detection of diabetic retinopathy, especially in regions with limited access to ophthalmologists. 

Released in November 2018, Google claimed the system was capable of identifying diabetic retinopathy with 90 percent accuracy in the testing laboratory.

Risks and harms 🛑

Google Health's diabetic retinopathy diagnosis system has been seen to potentially inaccurate, leading to misdiagnosis, pose privacy concerns regarding sensitive medical data, and over-reliance on technology that might diminish the role of professional medical judgement. 

Research, advocacy 🧮

• Beede E., et al. A Human-Centered Evaluation of a Deep Learning System Deployed in Clinics for the Detection of Diabetic Retinopathy

News, commentary, analysis 🗞️

• https://www.technologyreview.com/2020/04/27/1000658/google-medical-ai-accurate-lab-real-life-clinic-covid-diabetes-retina-disease/

• https://www.dailymail.co.uk/sciencetech/article-8285319/Googles-medical-AI-far-accurate-identifying-illness-clinics-laboratory.html

• https://www.forbes.com/sites/forbestechcouncil/2020/06/09/three-insights-from-googles-failed-field-test-to-use-ai-for-medical-diagnosis/

• https://www.newsweek.com/google-artificial-intelligence-deep-learning-health-research-thailand-clinics-diabetic-retinopathy-1500692

• https://techcrunch.com/2020/04/27/google-medical-researchers-humbled-when-ai-screening-tool-falls-short-in-real-life-testing/

• https://www.engadget.com/google-diabetic-blindness-reallife-tests-134057972.html