DeepSCEM: A User-Friendly Solution for Deep Learning-Based Image Segmentation in Cellular Electron Microscopy.

Date publication

septembre 2025

Journal

Biology of the cell

Auteurs

Membres identifiés du Cancéropôle Est :
Dr SCHULTZ Patrick

Tous les auteurs :
Meyer C, Hanss V, Baudrier E, Naegel B, Schultz P

Lien Pubmed

https://www.ncbi.nlm.nih.gov/pubmed/40888071

Résumé

Deep learning methods using convolutional neural networks are very effective for automatic image segmentation tasks with no exception for cellular electron micrographs. However, the lack of dedicated easy-to-use tools largely reduces the widespread use of these techniques. Here we present DeepSCEM, a straightforward tool for fast and efficient segmentation of cellular electron microscopy images using deep learning with a special focus on efficient and user-friendly generation and training of models for organelle segmentation.

Mots clés

cellular imaging, deep learning, electron microscopy, organelles, segmentation, software

Référence

Biol Cell. 2025 09;117(9):e70032