New Technology for Diagnosing Rare Conditions by Photograph May Foster Employment Discrimination?

As discussed in a recent article on PatientWorthy the FDNA has recently developed a new kind of algorithm which can diagnose rare genetic conditions such as Angelman syndromeWilliams syndrome, and Fragile X syndrome using merely a photograph of the patients face.

Its called DeepGestalt.

The company used 17,000 images to test this new technology and analyzed over 200 genetic diseases. The algorithm displayed a list of top 10 suggestions of diagnosis for each patient. For 91% of those in the study, the correct diagnosis was in this list of suggestions.

In three of FDNA’s trials, the technology out-performed clinical experts.

Researchers believe this technology could help standardize the way we describe physical attributes associated with rare conditions. Ultimately, this could improve future research and potentially even help identify new syndromes.

The previous article published on PatientWorthy discussed outsiders’ concerns with this new development. But the company has also been transparent about their own fears regarding their creation. 

The Fears

The problem comes down to the availability of the data FDNA uses to diagnose. Genomic data, used in genetic testing, is private. It’s concealed from public view. However, photographs are everywhere- from Facebook, to Instagram, to Twitter, to a simple google search.

FDNA is apprehensive that this could mean discrimination for those living with rare conditions.

“The technology could be open to abuse from employers who could use it to filter out workers who are less healthy and may take more time off or need greater support.”

Simply understanding the probability that someone may have a rare condition and therefore develop medical complications could deter employment of that individual.

We need to work to develop a way to eliminate this possibility.

Ultimately though, researchers are still hopeful about the positive impact this new technology may have. They believe that when combined with other forms of analysis, it could help patients receive diagnosis earlier. That often means improved treatment options and overall quality of life. We simply need a way to regulate this new technology.

You can read more about this take on this new algorithm here. You can also read the full report of this study, published in the Nature Medicine, here.

Share this post