February 10th, 2020
“Transmural Biotech presents the progress of its research on ‘deep learning’ to predict the gestational age of the fetus and improve the diagnosis of intrauterine growth restriction.”
Transmural Biotech presented the latest findings of the collaboration with BCNatal Fetal Medicine Research Center in diagnostic imaging using Artificial Intelligence and ‘deep learning’ in the Journal Club that the BCNatal research group organizes weekly, behind closed doors, to announce the progress of its project for the Automatic prediction of fetal age from brain ultrasound.
One of the reasons to be of Transmural Biotech is applying Artificial Intelligence (AI) to medicine. Their latest research in maternal-fetal medicine is a project for the automatic prediction of fetal age from brain ultrasounds through AI.
Dr Xavier P. Burgos-Artizzu, head of research at Transmural Biotech and specialist in developing diagnostic methods based on state-of-the-art image analysis technologies, explained this in detail in his presentation «Fetal age automatic prediction from brain ultrasound via deep learning” in front of BCNatal Fetal Medicine Research Center researchers.
A large bank of fetal brain images
The starting point of the research has been the development of the necessary infrastructure to create a unique maternal and fetal image bank, which classifies and organizes a large set of ultrasound imaging data of the fetal brain. Over the last two years, through collaboration with BCNatal Fetal Medicine Research Center, ultrasound images of several thousand patients have been collected.
Predict gestational age thanks to AI
Thanks to this large dataset of fetal brain ultrasound images, Transmural Biotech has been able to train a deep learning system capable of automatically detecting and recording images of the fetal brain. The objective is to estimate the gestational age taking advantage of the fetal brain structure.
The system was able to predict gestational age by an ultrasound image of the fetal brain alone, with a similar accuracy than that achieved by using all 4 classical fetal biometries from 3 different ultrasound images (Brain Biparietal Diameter, Head Circumference, Femur Length, Abdomen Circumference).
Work in progress
This progress has significantly improved the accuracy of the calculation of gestational age in third-trimester patients and in fetuses with restricted growth, the main focus of the project. The investigation, which is still underway, will allow us to know if, thanks to this system developed by Transmural Biotech and BCNatal, it is possible to predict the gestational age of the fetus automatically and more precisely than by conventional techniques (manual biometrics).
Towards a preventive and personalized medicine
As specialists in image analysis and processing technologies, Transmural Biotech enters the new technological era through the application of AI technologies, capable of seeing things that cannot be seen with the naked eye. The biotechnological solutions that it offers applied to the medical field allow a better diagnosis and treatment of patients, and lead us towards a preventive, personalized, non-invasive, fast and efficient medicine. The medicine of the future.
About the presentation
The presentation took place on February 10th in a classroom of the Hospital Clínic-Maternitat of Barcelona, within the framework of the Journal Club, weekly scientific event organized by BCNatal Fetal Medicine Research Center to see the evolution of its research projects.