As we know, it can be incredibly difficult to receive the proper diagnosis for a rare disease. In fact, it typically takes at least seven years and numerous doctor’s appointments. Because of this, the efforts of many medical professionals are aimed towards improving the diagnostic process for rare disorders. One effort in particular comes from the University of Bonn; researchers from this institution have developed an artificial intelligence to quicken and improve the diagnosis of rare, genetic disorders characterized by specific facial features.

Improving Diagnoses

A number of rare and genetic disorders share a similar symptom: abnormal facial features. One example is Noonan syndrome, in which affected individuals have distinctive features such as wide-set eyes and poor teeth alignment. As previously mentioned, diagnosis of these conditions can be difficult and take years. To combat this delay in diagnosis, researchers from the University of Bonn have developed an AI called GestaltMatcher.

Essentially, this AI is able to recognize different facial features associated with different genetic diseases, allowing it to suggest the proper diagnosis to doctors.

Armed with 17,560 photos of patients encompassing 1,115 rare diseases, the team of researchers has trained their software to analyze facial characteristics, calculate similarities, and link those faces to possible diagnoses, along with the clinical symptoms and genetic data of other patients. GestaltMatcher’s other capabilities include matching patients that it cannot find a diagnosis for to other undiagnosed patients with similar symptoms and recognizing diseases that were unknown to it before.

Looking Forward

This development is a huge one for the diagnosis of rare, genetic disorders. Using GestaltMatcher, patients will not only receive a diagnosis, but they’ll have a chance to receive treatment earlier. This is extremely important, as earlier intervention and treatment lead to better outcomes.

Not only will this technology help patients now, but it will spur future research and innovations as well. One of the researchers, Prof. Dr. Peter Krawitz, has pointed to retinal imaging and X-rays as the next sources of imagery to be used within their software. This could help an even larger patient population receive proper diagnoses.

Find the source article here.