The new Artificial Intelligence (AI) technology (Also called DeepGestalt) will be used to identify rare genetic diseases from the facial images of the patients. The AI technology outperformed the clinicians for the identification of many syndromes in 3 trials. This research could add a lot of value in personalized care, as per a study.
Almost 8% of the population can have genetic diseases and could be recognized through the facial features. For instance, the AI Technology could identify a disorder named Angelman syndrome; it is a disorder that affects the nervous system and it has features like widely spaced teeth, wide mouth, protruding tongue or even strabismus- a state where eyes look in different directions.
This research opens doors to a new technology which would help many people. But, as the facial images are easily available, many employers and payers could collect these images and gather information of their pre-existing conditions.
Gurovich along with his team used 17,000 facial images of the patients and trained DeepGestalt. Data of the patients was used to diagnose about 200 genetic syndromes for study. The tests conducted identified correct syndrome almost 91% of times.
Other tests were conducted for Noonan syndrome. It carries many distinctive features and health issues like heart problems. In these tests, there was 64% success. This technology can be useful in clinical settings. In this technology, a deep learning algorithm is applied to the photo of the face or the facial features and then, a list of probable syndromes is produced.
The tests were conducted with the patients who already were diagnosed with similar conditions. There was only a difficulty that the AI system’s performance will be very difficult to measure. The reason behind this is that there aren’t publicly available benchmarks yet.
A senior lecturer, Jorge Cardoso said that he finds this technology to be very interesting and he adds that this will help in reducing the burdens of healthcare systems. This will help to improve the health-care of the patients.
+61 3 9368 1377
Paul is a Founding Editor of conquesitor. Additionally, he is a content manager. As a full-fledged content writer and editor, he is devoted to writing news reports and bulletins about the everyday inventions, roll-outs, breakthroughs related to the business and technology field. Whenever free, he likes to spend time with kids and be like them. Thus, he visits orphanage home to help emerging bright futures.