Cardiotocography (CTG) and ultrasound are essential tools for assessing fetal well-being during pregnancy to detect potential complications early on. However, interpreting CTG recordings can be subjective and prone to misdiagnosis. This issue is exacerbated in low-resource facilities where access to skilled interpreters is even more limited.
Google researchers addressed this challenge by developing CTG-net, an AI model utilizing deep learning. It predicts fetal hypoxia, a critical condition of oxygen deprivation during labor. By analyzing time-series data of fetal heart rate and uterine contractions, the model provides an objective evaluation of CTG data, reducing reliance on expert interpretation. The model was trained using open-source CTG data from 552 patients and underwent testing in various clinical settings, including low-resource environments, proving its adaptability.
Google is currently exploring options to open-source the model. This groundbreaking advancement exemplifies the influence of AI on maternal healthcare and global health standards.
