NUCGEN3D: A synthetic framework for large-scale 3D nuclear segmentation with open-source training data and models
Emma Grandgirard, Theotime Dmitrasinovic, Corinne Barreau, Coralie Sengenès, Mathieu Serrurier, Ronan Sicre
The paper introduces NucGen3D, a synthetic framework for generating 3D nuclear segmentation datasets, and trains models that outperform existing methods on real-world data.
Computers in Biology and Medicine · Sep 3, 2026Peer-reviewedBioinformatics