Abstract
Ultrasonography (US) is the primary modality in prenatal assessment, as it provides insights into soft tissue characteristics, is widely available and cost-effective. However, variations in acoustic impedance across tissues, fetal motion, and probe motion degrade image quality, hindering objective assessment of fetal brain development. Brain atlases offer a standardized reference framework by aligning subjects within a common coordinate system, enabling consistent evaluation across populations. We propose ALRAU, a learning-based framework for constructing an artifact-reduced, time continuous fetal brain atlas from routine abdominal US data. Trained on over 300 volumes spanning 18–28 weeks of gestation, ALRAU achieves competitive image quality and pronounced anatomical definition of fine structures (Peak Signal-to-Noise Ratio: 24.6 dB). Age prediction on the constructed atlases confirms that age-discriminative anatomical features are preserved after registration. The mean absolute difference in predicted gestational age between test subjects and warped templates is 0.3 weeks. Warping the age-matched atlas to unseen subjects achieves robust alignment (Normalized Cross-Correlation: 0.81±0.07, Structural Similarity Index: 0.82±0.03) with perfect deformation invertibility, enabling reliable registration without anatomical distortion. This approach enables real-time, volumetric assessment of fetal brain development and supports informed prenatal decision-making. The model code will be made publicly available upon acceptance.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/ASMUS_053.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: https://papers.miccai.org/miccai-2026-sat/supp/ASMUS_053_supp.pdf
Link to Open Review
Open Review Page: https://openreview.net/forum?id=25b59LIvYZ
BibTex
@InProceedings{TisJoh_ALRAU_MICCAISAT2026,
author = { Tischer, Johannes AND Anzengruber, Stephan AND Mienkina, Martin AND Dorittke, Tim AND Binder, Julia AND Kasprian, Gregor AND Langs, Georg AND Licandro, Roxane},
title = { { ALRAU: Atlas Learning from Routine Antenatal Ultrasound } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17276},
month = {pending},
page = {pending}
}