Ultrasound AI's study validates advanced AI for predicting delivery timing using standard ultrasound images.
Key Details
- 1Ultrasound AI, in collaboration with University of Kentucky, published results in The Journal of Maternal-Fetal & Neonatal Medicine.
- 2The AI predicts time to delivery using only standard ultrasound images, not relying on clinical history or other risk factors.
- 3In a large cohort (over 2 million images, thousands of patients), the AI achieved an R² of 0.95 for term and 0.92 for all births.
- 4Continuous retraining improved AI's prediction of preterm births, with R² increasing from 0.48 (V1) to 0.72 (V4).
- 5Technology is scalable, non-invasive, and functions well across all trimesters and diverse patient populations.
Why It Matters
Accurate prediction of delivery timing, especially preterm births, remains a persistent challenge. This AI-driven approach, requiring only routine ultrasound images, could significantly improve pregnancy management and outcomes worldwide, particularly in under-resourced settings.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.