Cellpose 2.0: how to train your own model
2022/11/07 by Marius Pachitariu, Carsen Stringer · 69 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Engineering · #Architecture and Computational Design #Cell Image Analysis Techniques #Cellular Automata and Applications
paper · pdf · doi:10.1038/s41592-022-01663-4
openalex publication_date 2022/11/07 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/03
Abstract
Pretrained neural network models for biological segmentation can provide good out-of-the-box results for many image types. However, such models do not allow users to adapt the segmentation style to their specific needs and can perform suboptimally for test images that are very different from the training images. Here we introduce Cellpose 2.0, a new package that includes an ensemble of diverse pretrained models as well as a human-in-the-loop pipeline for rapid prototyping of new custom models. We show that models pretrained on the Cellpose dataset can be fine-tuned with only 500-1,000 user-annotated regions of interest (ROI) to perform nearly as well as models trained on entire datasets with up to 200,000 ROI. A human-in-the-loop approach further reduced the required user annotation to 100-200 ROI, while maintaining high-quality segmentations. We provide software tools such as an annotation graphical user interface, a model zoo and a human-in-the-loop pipeline to facilitate the adoption of Cellpose 2.0.
Citations
Cited by
- Creating and troubleshooting microscopy analysis workflows: Common challenges and common solutions
- Single cell resolved spatial immune repertoire unveils spatial heterogeneity of lymphoid aggregates in human immune disorders
- Cold-induced thermogenesis requires neutral-lipase-mediated intracellular lipolysis in brown adipocytes
- A multicellular actin star network underpins epithelial organization and connectivity
- Enabling neighbour labelling: using synthetic biology to explore how cells influence their neighbours
- Boundary constraints can determine pattern emergence
- Dynamics and plasticity of stem cells in the regenerating human colonic epithelium
- Live-cell imaging in the deep learning era
- Boundary-guided cell alignment drives mouse epiblast maturation
- Boundary geometry controls a topological defect transition that determines lumen nucleation in embryonic development
- GSDME–IL-18 pyroptotic axis prevents myosteatosis by expanding tissue-resident macrophages to promote muscle regeneration
- Practical guidelines for cell segmentation models under optical aberrations in microscopy
- GenCellAgent: Generalizable, Training-Free Cellular Image Segmentation via Large Language Model Agents
- Simultaneous, cortex-wide dynamics of up to 1 million neurons reveal unbounded scaling of dimensionality with neuron number
- Cell Instance Segmentation: The Devil Is in the Boundaries
- Cell State Transitions Beyond the Small-Noise Limit
- A hormetic transcriptional program coregulates invasion, proliferation and dormancy to define metastatic potential
- Optimizing the characterization and quantification of retinal ganglion cell somas in healthy and injured retinas using cellpose
- Automated X‐ray computed tomography‐based analysis of clast fabric in drill‐cores of glacial diamicts using deep learning models
- A rare PRIMER cell state in plant immunity
- Spatial multiomics of acute myocardial infarction reveals immune cell infiltration through the endocardium
- NANOG is repurposed after implantation to repress Sox2 and begin pluripotency extinction
- Highly adaptable deep-learning platform for automated detection and analysis of vesicle exocytosis
- Celldetective, an AI-enhanced image analysis tool for unraveling dynamic cell interactions
- Single-cell multi-omic atlas and morphogen screening informs midbrain and hindbrain organoid engineering
- Disuse‐induced muscle‐type specific alterations and adiponectin pathway response in male mice
- Doxorubicin-induced sinus node dysfunction associated with mitochondria and nuclear impairment in a mouse model
- RegVelo: Gene-regulatory-informed dynamics of single cells
- Dynamic Gene Expression Mitigates Mutational Escape in Lysis-Driven Bacteria Cancer Therapy
- MorphoNet 2.0: An innovative approach for qualitative assessment and segmentation curation of large-scale 3D time-lapse imaging datasets
- Tissue-like multicellular development triggered by mechanical compression in archaea
- The pigment cell : molecular, biological, and clinical aspects
- Mantis: high-throughput 4D imaging and analysis of the molecular and physical architecture of cells
- Midbrain node for context-specific vocalisation in fish
- Ndfip2 in TrkA-expressing sensory neurons regulates noxious mechanosensation through control of TrkA signaling and protein levels
- Sketchpose: Learning to Segment Cells with Partial Annotations
- Spatial transcriptomics reveals expression gradients in developing wheat inflorescences at cellular resolution
- A mother-machine microfluidic device for non-adherent mammalian cells reveals the population growth strategies
- Maybe you don't need a U-Net: convolutional feature upsampling for materials micrograph segmentation
- Fine-grained Multi-class Nuclei Segmentation with Molecular-empowered All-in-SAM Model
- SelfAdapt: Unsupervised Domain Adaptation of Cell Segmentation Models
- RedDino: A foundation model for red blood cell analysis
- ForSys: Non-invasive stress inference from time-lapse microscopy
- IAUNet: Instance-Aware U-Net
- An integrated single-nucleus and spatial transcriptomics atlas reveals the molecular landscape of the human hippocampus
- A COCO-Formatted Instance-Level Dataset for Plasmodium Falciparum Detection in Giemsa-Stained Blood Smears
- Muscle Regeneration Can Be Rescued in a Telomerase Deficient Zebrafish Model of Ageing by MMP Inhibition
- In vivo transcriptomic, functional, circuit-based, and translational analyses of enteric neurons
- Clonal-aggregative multicellularity tuned by salinity in a choanoflagellate
- Retinoic acid drives cell fate specification, maturation and retinal regionality in human retinal organoids
- Upregulation of an Epithelial miRNA Is Associated with Immune Evasion in Progressive Bronchial Premalignant Lesions
- Rho-ROCK liberates sequestered claudin for rapid de novo tight junction formation
- Leveraging AI for cell biology discovery
- Implementing distinct spatial proteogenomic technologies: opportunities, challenges, and key considerations
- Molecular recording of cellular protein kinase activity with chemical labeling
- Noradrenergic signaling controls Alzheimer’s disease pathology via activation of microglial β2 adrenergic receptors
- Radiation-induced amphiregulin drives tumour metastasis
- Chromosome mis-segregation triggers cell cycle arrest through a mechanosensitive nuclear envelope checkpoint
- Celldetective: an AI-enhanced image analysis tool for unraveling dynamic cell interactions
- The Four Color Theorem for Cell Instance Segmentation
- Multimodal profiling unveils a reversible breast basal-like cell state in AKT-inhibitor resistant tumours
- HIEGNet: A Heterogenous Graph Neural Network Including the Immune Environment in Glomeruli Classification
- Moment kernels: a simple and scalable approach for equivariance to rotations and reflections in deep convolutional networks
- The role of primary cilia in myoblast proliferation and cell cycle regulation during myogenesis
- Light-sheets and smart microscopy, an exciting future is dawning
- ROIsGAN: A Region Guided Generative Adversarial Framework for Murine Hippocampal Subregion Segmentation
- Optical microscopic imaging, manipulation, and analysis methods for morphogenesis research
- Time-resolved single-cell and spatial gene regulatory atlas of plants under pathogen attack
- UniBiomed: A Universal Foundation Model for Grounded Biomedical Image Interpretation
- The digital revolution in veterinary pathology
Related