Reproduce “Generalizable fetal brain biometry via adaptive artificial intelligence” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yijin Li, Mingxuan Liu, Yi Liao, Le Chang, Hongjia Yang, Kasidit Anmahapong, Xiaotian Hu, Ziang Wang and 19 more
BioTTA adapts a pretrained model to measure fetal-brain MRI biometry from unlabeled scans. It updates batch-normalization parameters while using geometric constraints and an age-matched atlas to guide landmark predictions. Across cross-sequence, cross-center, and cross-country tests, it lowered average measurement error versus source-only models. On FeTA, BioCNN achieved 1.460 mm MAE, a reported 47.1% reduction; it was not statistically distinguishable from supervised finetuning using ten labeled target volumes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.