AI for histology, in QuPath.

An open-source platform for histology image analysis. Use community models to identify cells and tissue structures, get measurements, and train and share models using your own annotations. Analysis and training run in your cloud account.

Contribute to HistoForge

Publish a module directly from QuPath. To discuss contributing a dataset or model, email histoforge.contact@gmail.com.

Explore HistoForge

Segment and classify cells

Lizard · Graham et al. · Image credit and overlay key

Segment and classify tissue structures

Da-Mi Kim · Image credit

Segment and classify tissue regions

RINGS · Salvi et al. · Image credit and overlay key

Run models within detected structures

Da-Mi Kim · Image credit and overlay key

Whole slide inference

CMU-1 · Image credit and annotation context

Calculate metrics

Da-Mi Kim · Image credit and measurement context

Train and fine-tune models

HistoForge · Interface capture

Create module workflows

Combine models and scripts Three example recipes. Leydig nucleus and cell models feed a measurement script. Tubule and lumen models feed a script that selects tubule cross-sections, followed by tubule measurements. Spermatogenesis uses the same tubule preparation, then runs a cell model inside selected cross-sections and counts cells per tubule. These are research examples, not a visual workflow editor or a list of available modules. Model Script Example recipes Leydig cells Tubules Spermatogenesis Leydig nuclei Leydig cells Cell countsand measurements Tubules Lumens Selectcross-sections Tubulemeasurements Cells withineach tubule Cell countsper tubule Selected cross-sections Reuses the selectedcross-sections above. Research examples · Models not yet released

Download community modules and share your own

HistoForge · Interface capture

Selected modules

On Hugging Face

For other modules, browse the histoforge tag on Hugging Face.

  • Da-Mi Kim Rat Tubule Staging (PAS)

    HoVer instance segmentation and 14-class stage or atrophy classification for PAS-stained rat seminiferous tubules.

    • dinov3-adapter-hovernet
    • histopathology
    • rat
    • testis
    • pas
    • instance-segmentation
    • image-classification
  • DeepLIIF Ki-67 Nuclei (IHC)

    Joint Ki-67-positive/negative nucleus segmentation, classification, counting, and positive-index aggregation.

    • dinov3-adapter-cellpose
    • pathology
    • immunohistochemistry
    • ki67
    • human
    • instance-segmentation
    • image-classification
  • DINOv3 Cellpose Starter

    DINOv3 Cellpose Starter for training on your own annotations.

    • starter
    • cellpose
    • histopathology
  • DINOv3 Dense Semantic Starter

    DINOv3 Dense Semantic Starter for training on your own annotations.

    • starter
    • semantic
    • histopathology
  • DINOv3 HoVer Starter

    DINOv3 HoVer Starter for training on your own annotations.

    • starter
    • hovernet
    • histopathology
  • H&E-12 Cellpose Starter

    H&E-12 Cellpose Starter for training on your own annotations.

    • starter
    • cellpose
    • histopathology
  • H&E-12 Dense Semantic Starter

    H&E-12 Dense Semantic Starter for training on your own annotations.

    • starter
    • semantic
    • histopathology
  • H&E-12 HoVer Starter

    H&E-12 HoVer Starter for training on your own annotations.

    • starter
    • hovernet
    • histopathology
  • IGNITE NSCLC Tissue Segmentation (H&E)

    Tissue-compartment segmentation and area measurements in H&E-stained non-small cell lung cancer.

    • histopathology
    • semantic-segmentation
    • dinov3
  • Lizard Colon Nuclei (H&E)

    Segment and classify six nucleus types in colon H&E images.

    • dinov3-adapter-cellpose
    • pathology
    • human
    • colon
    • instance-segmentation
    • image-classification
  • NuCLS Breast Nuclei (H&E)

    Nucleus instance segmentation and seven-class classification in human breast-carcinoma H&E fields.

    • dinov3-adapter-cellpose
    • pathology
    • human
    • breast
    • he
    • instance-segmentation
    • image-classification
  • PUMA Melanoma Nuclei (H&E)

    Melanoma nucleus segmentation, ten-class classification and nucleus counts.

    • dinov3-adapter-cellpose
    • pathology
    • human
    • melanoma
    • instance-segmentation
    • image-classification
  • RINGS Prostate Gland and Tumor Regions (H&E)

    Independent dense-semantic gland and tumour-region segmentation for H&E prostate fields.

    • dinov3-adapter-semantic
    • pathology
    • prostate
    • human
    • he
    • semantic-segmentation
  • Kidney glomeruli

    Glomerulus segmentation and morphometric measurements in kidney tissue.

    • Orbit
    • Stritt
  • Glomerulosclerosis

    Glomerulus segmentation and classification by sclerosis status.

    • AIDPATH
    • HuBMAP
  • Liver zonation

    Portal-field and central-vein detection for spatial analysis of liver zones.

    • PF/CV
  • Liver lesions and mitosis

    Necrosis-region analysis and detection of mitotic and injured cells.

    • TG-GATEs
  • Gleason grading

    Gleason-pattern analysis in prostate tissue.

    • Arvaniti
    • SICAPv2
    • PANDA
  • Breast tissue regions

    Tissue-compartment segmentation in breast cancer images.

    • BCSS
  • Lymph-node metastasis

    Tumour-region detection and segmentation in lymph-node sections.

    • CAMELYON16
    • CAMELYON17
  • Colon tissue regions

    Tissue-region classification and segmentation in colorectal images.

    • HunCRC
    • EBHI-Seg
  • Cross-species mitosis

    Mitotic-figure detection and counting in human and animal tissue.

    • MIDOG++
    • CCMCT
    • CMC
  • Tumour-infiltrating lymphocytes

    Lymphocyte detection and density measurements within tumour and stroma regions.

    • TiGER
  • PD-L1 scoring

    Tissue and cell analysis for tumour proportion and combined positive scores (TPS/CPS).

    • IGNITE
  • Canine skin tumours

    Tissue-region segmentation and tumour classification in canine skin.

    • CATCH
  • Bronchoalveolar lavage cytology

    Cell detection and assessment of hemosiderin in bronchoalveolar lavage samples.

    • P-Hem
  • Ovarian follicles

    Follicle detection, classification and counting in ovarian tissue.

    • Blot
    • MOTHER
  • Pan-tissue nuclei

    Nucleus segmentation and classification across tissue types.

    • PanNuke
  • Colon nuclei

    Nucleus segmentation and classification into six cell types in colon tissue.

    • Lizard
    • CoNIC
  • Multi-organ nuclei

    Nucleus segmentation and classification in breast, kidney, lung and prostate tissue.

    • MoNuSAC
  • CoNSeP colon nuclei

    Nucleus segmentation and classification in colorectal cancer tissue.

    • CoNSeP
  • Melanoma nuclei

    Nucleus segmentation and cell-type classification in melanoma tissue.

    • PUMA
  • Gastric nuclei

    Nucleus segmentation and classification in gastric tissue.

    • GlySAC
  • Canine and feline lymphoma

    Nucleus analysis in canine and feline lymphoma images.

    • FH-OOe