“Ray Hammond delivered a great keynote. He is a superb speaker.” Google
“Ray gave us an incredible, upbeat glimpse into the future.” Mastercard
“Ray Hammond is one of the best public speakers we have ever seen.” GE Corporation, Europe
“A really stunning keynote speech.” Microsoft
“Ray’s speech was uplifting, inspiring and visionary.” Vodafone
“Ray stimulated and inspired our North American sales team at the end of a very long day.” Ericsson Networks
“As a speaker, Ray Hammond has Mojo!” National Property Council of Australia.
“Ray Hammond was fabulous. He bent our minds.” Intel Corporation
“Ray – thank you for your terrific contribution to our Business Leadership Forum.” McKinsey & Co
“Ray was awesome at our 4th Annual Seminar in Montego Bay.” Ministry of Tourism, Jamaica
“Ray was brilliant at our client conference in Cape Town.” Telkom SA

Latest Tweets

YouTube Subscribe

Downloadable PDFs

Glimpses Of The Future

Selfies That Can Detect Clogged Arteries

 A new algorithm has been developed to detect coronary artery disease solely from patient facial photos. The proof-of-concept, published in the European Heart Journal, needs more refinement before it becomes a useful clinical tool but independent experts also suggest there are profound ethical considerations that need to be resolved before a system like this can be deployed.

Alopecia, xanthelasmata (yellowing on the eyelids), and arcus corneae (an opaque ring around the cornea), are among several facial biomarkers to indicate a person may be suffering poor cardiovascular health. A team of researchers from China has now developed a deep learning algorithm that can study just four photos of an individual to determine a person’s risk of coronary artery disease.

Over two years, between 2017 and 2019, the researchers recruited 5,796 patients who had presented at a hospital to undergo heart imaging procedures. Four photos were taken of each patient – two side profiles, one frontal, and one looking down at the top of the head.

A deep learning algorithm was then trained to study these four images and assess a person’s heart disease risk. The results, validated in a further 1,000 patients, saw the algorithm correctly detecting heart disease in 80 percent of the cohort. Additionally, the algorithm could accurately detect patients without coronary artery disease 61 percent of the time.

The ethical resolves around who owns the data and must permission for analysis be sought every time?

Read full article

No Fields Found.