AI Facial Recognition Could Help Identify Genetic Disorders

Published on January 17, 2019

An AI facial recognition software that can detect genetic disorders just by looking at a picture of your face is coming. But will it be as simple as it sounds?

Facial recognition softwares are all too common these days. AI softwares like these can be used to do things like unlock your phone, go through immigration, or even rent a car. There has been a lot of skepticism surrounding the potential harmful impacts of AI (ahem, Black Mirror). The risks involved, though, don’t outweigh the benefits as the software has the potential to make our lives easier, healthier, and more stress-free. Use a self-learning thermostat and tell me you’re against AI. I dare you.


Using facial recognition to find genetic disorders means that scanning your body for potentially life-threatening diseases could be part of your normal checkup. A US-based company called FDNA published a study in a journal called nature medicine about their new software called DeepGestalt. The software has been testing in clinical trials to find genetic disorders through an AI facial recognition technology. The technology has, apparently, done well in clinical trials and even outperformed its human counterparts in being able to identify quite a few syndromes.

The study shows that an algorithm can determine what percentage of a population has key facial features that coincide with specific genetic disorders. It’s able to study these features and match them against its known database to determine whether the patient is at-risk. One of the disorders, Noonan Syndrome, carries a very specific range of facial features that contribute to the disorder. Noonan Syndrome can contribute to things like heart disease and prevents normal development.


As with any technology, there are some significant downsides to the possibility of using AI facial recognition in healthcare. You may have heard rumor that DNA testing companies like 23 and Me are harvesting sensitive data. Why? To sell it to insurance companies. These insurance companies can then use that data to discriminate against entire demographics. They would do that based on algorithmic data that matches their DNA to pre-existing conditions. The same issue arises with AI facial recognition algorithms being able to determine pre-existing conditions based simply on a photo.

Ideally, the law would be able to catch up to things like this. It would prevent insurance companies from being able to discriminate in such ways. The government is a bit busy, though, so it’s unlikely we’ll see the law catch up to advancements in technology out of the blue. If used for good though, AI facial recognition softwares like DeepGestalt could provide groundbreaking innovations for the medical industry.


Julia Sachs is a former Managing Editor at Grit Daily. She covers technology, social media and disinformation. She is based in Utah and before the pandemic she liked to travel.

Read more

More GD News