Abstract
The training of fetal ultrasound sonographers is limited by a shortage of experienced trainers. This questions whether an AI system can be developed to provide scalable ultrasound training support. However, existing AI-assisted fetal ultrasound models typically focus on image interpretation or quality assessment, but do not translate assessment scoring into acquisition guidance. We introduce A3F, a multi-task Anatomy-Aware Actionable Feedback framework that directly guides a trainee on image magnification, gain, anatomy centering, and acoustic shadow without an explicit intermediate image interpretation stage. We also establish a design principle for actionable ultrasound feedback: clinically deployable guidance should be succinct and action-focused to minimise cognitive load on a sonographer. Results show that decoupling actionable feedback categories into independent prediction heads enables logically consistent prediction of co-occurring instructions, without compromising real-time inference speed. Ablation study demonstrates that anatomical context injection and anatomy-specific augmentations further strengthen model performance. These findings establish the feasibility of directly learning clinically meaningful actionable feedback from fetal ultrasound images, and motivate future user studies to evaluate its impact on sonographer training. We make the code publicly available, alongside a lightweight annotation tool tailored for fast annotation used in the research. Both are available at: https://github.com/angela-feixue-wang/ A3F.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/HAIC26_017.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=HwcOdzHFjq
BibTex
@InProceedings{WanAng_A3F_MICCAISAT2026,
author = { Wang, Angela Feixue AND Lamdouar, Hala AND Lander, Jayne AND Papageorghiou, Aris T. AND Noble, J. Alison},
title = { { A3F: Anatomy-Aware Actionable Feedback for Fetal Ultrasound } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17279},
month = {pending},
page = {pending}
}