New AI tool developed by researchers helps in diagnosing rare disorders in kids by scanning faces

The fast progression of AI has made the world fall into a perplexed state. People are feeling a myriad of emotions towards it. According to the results published by Oberlo, 62% of people do not mind AI assistance, while the remaining are wary of it. Some are scared that it will take over their role and functions, while some think its assistance will make the world a better place. Moreover, the medical community has been using AI in many areas for research, diagnosis, and practical application. Through the aid of AI, they have been able to make many breakthroughs. One such breakthrough has been achieved by researchers at Radboud University Medical Center, who have created an AI model that can diagnose syndromes in kids by scanning their faces, as reported by AD.

This breakthrough can solve a massive complication faced by parents when it comes to the diagnosis of their children. There are times when it takes years to find out what syndromes are afflicting the children. It leads to delays in proper treatment, which makes children more weak and susceptible to other issues. Lex Dingemans, a Ph.D. student involved in the project, was happy with the performance displayed by the AI model, “We gave the computer the task: recognize these syndromes using the face and medical record of a group of children with forty different syndromes. That went amazingly well." They tested the AI model on 40 specimens and it was able to give the correct diagnosis in 37 cases. The model also has a feature that can find the link present between syndromes and particular genetic abnormalities.

This feature will help doctors who face problems in treating a patient due to the lack of case studies and research behind similarities. The feature of identifying similar abnormalities will help them formulate a record of case histories that they can use in their course of action. The doctors can now connect a case from Adelaide in Australia with a case in Rotterdam, having similar abnormalities as pointed out by the AI.

Clinical geneticist Bert De Vries shares why there is often a delay in correct diagnosis and selecting a path of treatment, “New syndromes are still emerging, and sometimes only a few dozen cases are known worldwide.” It leads to uncertainty for a long time. Doctors and parents are not fully sure due to the lack of a database, whether the syndrome diagnosed is correct and faulty treatment could bring more problems.

This model is helpful, especially in the case of rare syndromes. De Vries explained, “Everyone is capable of recognizing children with Down syndrome. They look more like each other than their brothers or sisters. But there are countless other rare disorders that are much less common." It often takes years to diagnose these syndromes. Apart from mild to moderate intellectual issues, these syndromes also result in physiological problems in other organs like the heart and kidneys. Correct diagnosis helps in quick preventive measures for the kids in question. It also helps in making the quality of life better for such kids. The researchers are hoping to roll out the AI model for use as soon as possible.