Machine Learning Tool Detects Genetic Syndromes in Children

Children’s National Hospital researchers developed a machine learning tool that offers rapid genetic screening, potentially accelerating the diagnosis of genetic syndromes in children.

The deep learning technology was trained with data from 2,800 pediatric patients from 28 countries. Additionally, the technology also considers variables related to sex, age, racial, and ethnic backgrounds.

“We built a software device to increase access to care and a machine learning technology to identify the disease patterns not immediately obvious to the human eye or intuition, and to help physicians non-specialized in genetics,” principal investigator in the Sheikh Zayed Institute for Pediatric Surgical Innovation at Children’s National Hospital and senior author of the study Marius George Linguraru, DPhil, MA, MS, said in a press release.

“This technological innovation can help children without access to specialized clinics, which are unavailable in most of the world. Ultimately, it can help reduce health inequality in under-resourced societies.”

The machine learning technology can indicate the presence of a genetic syndrome from a facial photograph captured at a point-of-care, such as in pediatrician offices, maternity wards, and general practitioner clinics.

“Unlike other technologies, the strength of this program is distinguishing ‘normal’ from ‘not-normal,’ which makes it an effective screening tool in the hands of community caregivers,” said Marshall L. Summar, MD, director of the Rare Disease Institute at Children’s National. 

“This can substantially accelerate the time to diagnosis by providing a robust indicator for patients that need further workup. This first step is often the greatest barrier to moving towards a diagnosis. Once a patient is in the workup system, then the likelihood of diagnosis (by many means) is significantly increased.”

Every year, millions of children are born with genetic disorders, including Down syndrome, Williams-Beuren syndrome, and Noonan syndrome. Most children with genetic syndromes live in areas with limited resources and access to genetic services, creating health disparities.

Additionally, there is an insufficient number of specialists to help identify genetic syndromes early in life when preventive care can save lives, especially in areas of low incomes, limited resources, and isolated communities.

According to the researchers, introducing the technology to pediatricians, neonatologists, and family physicians for pediatric evaluation, especially in areas with limited access to specialized care, could be a step forward in democratizing health resources for genetic screening.

The technology was trained using 2,8000 retrospective facial photographs of children with and without genetic syndromes from 28 countries, including Argentina, Australia, Brazil, China, France, Morocco, Nigeria, Paraguay, Thailand, and the US. The deep learning method was designed to account for the normal variations in face appearances among diverse populations.

“Facial appearance is influenced by the race and ethnicity of the patients. The large variety of conditions and the diversity of populations are impacting the early identification of these conditions due to the lack of data that can serve as a point of reference,” said Linguraru. “Racial and ethnic disparities still exist in genetic syndrome survival even in some of the most common and best-studied conditions.”

The machine learning tool was trained on the available dataset. As more data from underrepresented groups becomes available, researchers will adapt the model to localize phenotypical variations with more specific demographic groups.

Additionally, researchers envision the utility of this technology in new screenings.

“There are approximately 140 million newborns every year worldwide of which eight million are born with a serious birth defect of genetic or partially genetic origin, many of which are discovered late,” said Linguraru.