Cellpose3: one-click image restoration for improved cellular segmentation

By C. Stringer, Marius Pachitariu. bioRxiv, 2024

377 citations recorded by the source corpus.

Generalist methods for cellular segmentation have good out-of-the-box performance on a variety of image types. However, existing methods struggle for images that are degraded by noise, blurred or undersampled, all of which are common in microscopy. We focused the development of Cellpose3 on addressing these cases, and here we demonstrate substantial out-of-the-box gains in segmentation and image quality for noisy, blurry or undersampled images. Unlike previous approaches, which train models to restore pixel values, we trained Cellpose3 to output images that are well-segmented by a generalist segmentation model, while maintaining perceptual similarity to the target images. Furthermore, we trained the restoration models on a large, varied collection of datasets, thus ensuring good generalization to user images. We provide these tools as “one-click” buttons inside the graphical interface of Cellpose as well as in the Cellpose API.

Subfields: Biophysics, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/f396860627bf43ffe714d0ae749ac588491e5483

Citations in the corpus

• Single-cell landscape of peripheral and tumor-infiltrating immune cells in HPV-negative HNSCC (2025, unknown): “CellPose 3.0 was used for image segmentation, 62 and regions were classified as whole slide, intra-tumoral, or peri-tumoral by a pathologist.”
• The Single-Cell Landscape of Peripheral and Tumor-infiltrating Immune Cells in HPV- HNSCC (2025, unknown): “47 Segmentation objects were categorized into three architectural features—whole slide, intra-tumoral, and peri-tumoral—based on a pathologist’s assessment of the tissue.”