What To Know
- patent application states PAI's software has been applied to real-world data obtained from a clinical context and has been shown to consistently produce a positive predictive value (the AI predicts a preterm birth, and the mother then delivers preterm) above 90% and a negative predictive value (the AI predicts a normal birth, and then the mother delivers the baby at full term) above 90%.
- After he and his wife experienced nine miscarriages and the happy arrival of four daughters over time, he knew he wanted to “make a huge dent in every serious affliction that only impacts women.
Ultrasound AI, Inc., an artificial intelligence company dedicated to improving women’s health, today announces that they were awarded their first patent on International Women’s Day for the software powering their groundbreaking predictive diagnostic medical technology, Preterm AI™ (PAI). PAI combines the power of ultrasound with artificial intelligence to predict preterm birth.
The U.S. patent application states PAI’s software has been applied to real-world data obtained from a clinical context and has been shown to consistently produce a positive predictive value (the AI predicts a preterm birth, and the mother then delivers preterm) above 90% and a negative predictive value (the AI predicts a normal birth, and then the mother delivers the baby at full term) above 90%.
“This technology has the potential to revolutionize obstetric care as well as accelerate research to develop new interventions to prevent both premature births and miscarriages,” said Dr. Garrett K. Lam, FACOG, Chief Medical Officer at Ultrasound AI.
Preterm birth is when a baby is born before 37 weeks of pregnancy. In 2020, preterm birth affected one in 10 infants born in the United States.1 Preterm birth complications are the leading cause of death among children under five years of age, responsible for approximately one million deaths annually.2 Premature births often result in emotional trauma for families, prolonged stays in specialty care nurseries, and profound health consequences for preterm infants.3 Many women who deliver prematurely have no known risk factors or early symptoms of complications.4
“Doctors assume 280 days until birth, and they may be able to label someone as ‘at risk,’ but that isn’t necessarily accurate. The ability to accurately predict if a baby will be born early and how early are non-existent using ultrasound imaging. The current blood tests are limited in use and accuracy, and can be expensive,” explains Robert Bunn, President and Founder of Ultrasound AI.
Bunn is the pioneering expertise behind this discovery. After he and his wife experienced nine miscarriages and the happy arrival of four daughters over time, he knew he wanted to “make a huge dent in every serious affliction that only impacts women.” He used AI to analyze the data from 400,000+ de-identified ultrasound images. The AI learned patterns from those images and was eventually able to recognize patterns in images and associate them with birth outcomes.
“There’s a lot of noise in the images in ultrasound, so humans have difficulty seeing the image and understanding it. Our AI software is trained to do what we can’t and see things that we can’t see. Most people wouldn’t even consider this possible,” shared Bunn.
Now, after four years of cutting-edge research and development, this revolutionary new AI software shows the promise to change the lives of countless families around the world. PAI has the power to give providers and patients timely, accurate information to make better-informed decisions and improve the wellbeing of babies and mothers.
