Search papers, labs, and topics across Lattice.
This study introduces a foundation model-driven framework for selecting key anatomical frames from blind-sweep ultrasound videos to estimate fetal birth weight (FBW) shortly before delivery. By leveraging a vision-language model for anatomy-guided frame selection and a redundancy-aware feature compression module, the method achieves significant accuracy in FBW regression without relying on operator expertise. The approach was validated on data from 839 patients, yielding a mean absolute error of 161.3 g and outperforming traditional estimation methods, demonstrating its potential for operator-independent assessments in low-resource settings.
Achieving a mean absolute error of just 161.3 g, this method revolutionizes fetal birth weight estimation by eliminating the need for expert operators in ultrasound assessments.
Accurate fetal birth weight (FBW) estimation shortly before delivery is clinically valuable yet challenging due to its reliance on operator expertise, particularly in low-resource settings. To reduce this reliance, we study near-term birth-weight regression from blind-sweep ultrasound (US) videos acquired within 48 hours prior to delivery, with post-delivery weighing as ground truth. Accordingly, we propose a foundation model-driven key anatomy frame selection framework that enables accurate FBW regression despite the absence of plane constraints in blind sweeps. Our highlights are as follows: (1) We believe this is the first work to estimate FBW using blind-sweep US videos, enabling operator-independent assessment. (2) An Anatomy-Guided Frame Selection module equipped with a vision-language foundation model is proposed for keyframe collection in unconstrained sweeps. (3) A Redundancy-Aware Feature Compression module is designed to compress frame features while preserving task-relevant information, alleviating temporal redundancy. Extensively validated on prospectively collected data from 839 patients, our method achieves an MAE of 161.3 g, with 90.23% and 100% of cases falling within 10% and 15% absolute percentage error, outperforming typical Hadlock estimation and strong competitors. Codes are available at https://github.com/ouleoule/BlindSweep-EBW.
