Researchers at the University of Ottawa’s Faculty of Medicine have pioneered the use of an Artificial Intelligence-based deep-learning model to assist in the accurate and rapid reading of ultrasound images.
The study’s purpose was to demonstrate deep-learning architecture that could be used to quickly and reliably identify cystic hygroma in the first trimester of ultrasound scans. Cystic Hygroma, an embryonic condition, causes the lymphatic system to develop abnormally. It is a rare condition that can cause fluid swelling of the neck and head.
The birth defect can usually be diagnosed prenatally with an ultrasound. But Dr. Walker, cofounder of The Ottawa Hospital’s OMNI Research Group, (Obstetrics, Maternal and Newborn Investigations), and his research team wanted to determine if AI-driven pattern detection could help.
“What we showed was that in the field ultrasound, we were able to use similar tools for image classifications and identification with high sensitivity and specificity,” Dr. Walker said. He believes the approach could also be applied to other fetal anomalies commonly identified using ultrasonography.