AnatoProto is a lightweight sequence‑level framework that adapts a frozen BiomedCLIP encoder for detecting the fetal abdominal circumference standard plane in low‑cost obstetric blind sweeps. It incorporates anatomy‑weighted spatial pooling, a within‑case prototype loss, a three‑stage cascade refinement, and a hybrid prediction head to address the highly imbalanced, short‑segment nature of the task. On the ACOUSLIC‑AI benchmark, AnatoProto achieves a test F1 of 67.72, surpassing the best foundation‑model baseline by 13.20 F1 and the best video temporal‑action‑detection baseline by 15.76 F1.
arXiv:2608. 04766v1 Announce Type: cross Abstract: A large number of infants with congenital anomalies are born each year globally, especially in areas with underdeveloped medical resources.
arXiv:2609.19230v1 Announce Type: new Abstract: Ultrasound is the most widely deployed imaging modality worldwide, yet clinical AI remains fragmented into narrow single-task models that fail when dev...
This study introduces a pipeline that enhances ultrasound plane pose estimation for fetal brain imaging by providing continuous, real‑time proximity feedback to standard planes (SPs). It employs a semi‑supervised segmentation model achieving high mIoU scores on both SPs and non‑SPs, and integrates a classification step to filter out frames without the fetal brain. The system, validated on an NVIDIA Clara AGX edge device, runs at 39 Hz and has been tested on real scan videos from 17 sonographers, demonstrating its practical viability for clinical use.
arXiv:2606. 29586v1 Announce Type: cross Abstract: Vision-language foundation models have shown strong potential in medical image analysis.
arXiv:2606. 11106v1 Announce Type: cross Abstract: A global shortage of trained sonographers limits prenatal ultrasound screening in low- and middle-income countries, where over half of pregnant women receive no skilled sonography.
arXiv:2608.00231v2 Announce Type: replace Abstract: Volumetric CT vision-language pretraining learns 3D representations from scan-report pairs, but global and anatomy-aware objectives supervise only...
arXiv:2608.30844v1 Announce Type: cross Abstract: Interactive lesion segmentation in whole-body PET/CT requires a model to provide a strong initial prediction while also responding efficiently to spa...
arXiv:2608.28455v1 Announce Type: new Abstract: Contrastive vision-language learning uses paired chest CT volumes and radiology reports to learn abnormality classifiers without manually annotated lab...
arXiv:2608. 00195v1 Announce Type: cross Abstract: High-resolution 3D segmentation of hip and shoulder anatomy from CT and MRI is essential for surgical planning, yet frozen segmentation models often fail under domain shift.
arXiv:2608. 14763v1 Announce Type: cross Abstract: Assessment of ventriculomegaly (VM) on fetal brain ultrasound relies primarily on measuring lateral ventricular atrial width on standard planes, which is operator-dependent and may not fully reflect the overall ventricular enlargement.
arXiv:2607. 27154v1 Announce Type: cross Abstract: CT vision-language foundation models have demonstrated promising performance across downstream tasks, but are typically trained with whole-volume representations that dilute fine-grained anatomical signals.
The paper presents the first systematic evaluation of out‑of‑distribution generalization for congenital heart disease (CHD) segmentation, using the ImageCHD cohort as a held‑out target. It compares several segmentation architectures under different training regimes, showing that in‑distribution performance is a poor predictor of cross‑cohort robustness: nnU‑Net drops from 0.77 to 0.51 Dice, while SwinUNETR maintains higher performance at 0.67 Dice. Limited target‑domain adaptation with only 11 labeled ImageCHD cases boosts all SwinUNETR variants above 0.76 Dice, highlighting the importance of explicit cross‑dataset testing.