3DCPath-datasets: A compilation of publicly available datasets in 3D pathology

3DCPath-datasets is a compilation of publicly available datasets in 3D Computational Pathology (3D CPath), spanning morphological data (3D tissue images) and molecular 3D data. It is organized into three main categories: volumetric morphological/imaging data, volumetric genomic data, and combined volumetric morphological and genomic data. The goal of this repository is to accelerate research in 3D CPath and provide the community with a centralized place to access 3D pathology datasets for future studies.

[!NOTE]
Contributions are welcome! Please check the contributions section below!

Updates

• 08/25/26: 3DCPath-datasets is now live and will be continuously updated as new datasets become available.

Volumetric morphological/imaging data

  Added
  Venue / Date
  Dataset
  Title
  Modality
  Resolution
  Organ
  Volumes
  Size




  08/25/26
  Nature Biomedical Engineering (2026)

    TRICARE

    [Paper]

  Deep-learning triage of three-dimensional pathology datasets for comprehensive and efficient pathologist assessments
  Open-Top Light-Sheet Microscopy (OTLS)
   Prostate: ~1 µm lateral, ~3.5–4 µm axial; Esophagus: ~0.6 µm lateral, ~2.75–3 µm axial
  Prostate cancer biopsies and endoscopic biopsies of Barrett’s esophagus
   Esophageal: 29 patients (125 specimens); Prostate: 59 patients (59 specimens)
  191.3GB


  08/25/26
  Cell (2026)

    ULTRA

    [Paper]

  Ultrarapid deep 3D histology enables intraoperative mapping of glioma infiltration
  Volumetric stimulated Raman scattering (SRS) lipid-channel images using the ULTRA pipeline
   Lateral resolution: 0.414 μm/pixel, axial resolution: 1 μm
  Three major glioma subtypes: oligodendroglioma, astrocytoma, and glioblastoma
  3 volumes with over a depth of 1 mm
  9.4GB


  07/15/26
  Royal Society Interface (2026)

    VISTACT

    [Paper]

  VIrtual histological STAining of micro-Computed Tomography (VISTACT)
  Synchtron-based phase-contrast microCT and serial H&E and EvG-stained histological sections
   Human lung microCT scans: 1.63 μm/voxel ; Mouse lung microCT scan: 0.88 μm/voxel
  Human lung tissue from patients with pulmonary hypertension and mice heart-lung tissue
  10 from human tissue (4 patients), and 1 from mice tissue
  869GB


08/25/26
Cell Press Blue (2026)

  CODA matrices 1

  CODA matrices 2

  [Paper]

Quantitative 3D histology reveals localized immune remodeling during early pancreatic cancer progression
CODA-derived 3D tissue and cell matrices from serial H&E histology, with IHC-derived immune-cell information for selected samples
Released matrices: 12 × 12 × 12 µm³/voxel; underlying CODA tissue segmentation generated at 2 µm/pixel
Human pancreas
48 samples / 48 volumes; two samples additionally have IHC-enhanced matrix versions
99.7 GB


  07/15/26
  Nature Communications (2025)

    Holotomography

    [Paper]

  Revealing 3D microanatomical structures of unlabeled thick cancer tissues using holotomography and virtual H&E staining
  Holotomography and H&E-virtually-stained images
  Lateral resolution: 156 nm; Axial resolution: 1.07 μm
  Human colon cancer samples
  N/A
  24.74GB


  08/25/26
  eLife (2024)

    OT-TP-LSM

    [Paper]

  Open-top Bessel beam two-photon light sheet microscopy for three-dimensional pathology
  Open-top Bessel-beam two-photon light-sheet microscopy
   Isotropic 0.9 µm/voxel
  Human skin—basal-cell carcinoma; pancreas—normal, IPMN/premalignant, and cancer; prostate—benign and cancer
   7 human tissue volumes: 2 skin BCC, 3 pancreas (normal, IPMN, PDAC), and 2 prostate (benign, adenocarcinoma)
  11.10 GB


  07/15/26
  Cancer Research(March 2023)

    PCa_Bx_3Dpathology

    [Paper]

  3D pathology of prostate biopsies with biochemical recurrence outcomes: raw H&E-analog datasets and image translation-assisted segmentation in 3D (ITAS3D) datasets
  OTLS
  0.88 μm/pixel
  Human prostate
  118 (50 patients)
  3.8TB


  07/15/26
  MICCAI challenge(2023)

    ACROBAT


  AutomatiC Registration Of Breast cAncer Tissue (ACROBAT)
  H&E and IHC (ER, PGR, HER2, KI67)
   Multiple. Validation conducted at 1.25x
  Human breast cancer
  750 (1 H&E and 1-4 matched routine IHC)
  n/a


08/25/26
European Journal of Cancer (2023)

  3D melanoma histology model

  [Paper]

A 3-dimensional histology computer model of malignant melanoma and its implications for digital pathology
Serial H&E histological sections, reconstructed into a 3D volume using TrakEM2/Fiji
Original scans: 0.25 µm/pixel at 40×; alignment images downsampled to 1 µm/pixel; 3 µm section thickness
Human cutaneous malignant melanoma
1 volume (66 serial H&E sections)
732.7MB


  07/15/26
  Optica(December 2022)

    Data

    [Paper]

  Label- and slide-free tissue histology using 3D epi-mode quantitative phase imaging and virtual hematoxylin and eosin staining
  3D epi-mode quantitative phase imaging
  Lateral resolution of 0.6µm and cross-sectional/axial resolution of 3.5 µm
  Mouse liver, rat gliosarcoma, and human gliomas
  Mouse liver: 8,  rat gliosarcoma: 14, human gliomas: 5 
  26 GB


  07/15/26
  Cell(2022)

    3DCRC

    [Paper]

  Multiplexed 3D atlas of state transitions and immune interaction in colorectal cancer
  H&E and t‑CyCIF serial histological sections
  Tissue sections with 5μm thickness and variable spacing. Images scanned at 20x 
  Colorectal cancer sample
  1 (47 serial histological sections)
  2.1 TB


  07/15/26
  PNAS(2022)

    3d virtual histology of the human hippocampus

    [Paper]

  Three-dimensional virtual histology of the human hippocampus based on phase-contrast computed-tomography
  Phase-contrast microCT
  160 nm/px and 50 nm/px
  Human hippocampus tissue
  N/A
  64.8 GB


  07/15/26
  eLife(2021)

    3D virtual Histopathology

    [Paper]

  3D virtual histopathology of cardiac tissue from Covid-19 patients based on phase-contrast X-ray tomography
  Phase-contrast microCT (Histomography scanner)
  1.3µm
  Cardiac Tissue from Covid-19 Patients
  6
  175.4 GB


  07/15/26
  eLife(2020)

    3D virtual pathohistology

    [Paper]

  3D virtual pathohistology of lung tissue from Covid-19 patients
  Phase-contrast microCT (Histomography scanner)
  1.3µm, 650nm, and 167nm
  Human lung tissue
  6 Covid-19 patients and one control tissue (healthy lung)
  568.4GB

08/25/26
Bioinformatics (2018)

3D histology reconstruction benchmark

[Paper]

Comparative analysis of tissue reconstruction algorithms for 3D histology
Serial H&E whole-slide images, tissue masks and manually annotated landmarks
0.46 µm/pixel, scanned at 20×; 5 µm section spacing
Murine prostate and murine liver
2 volumes: prostate = 260 sections; liver = 47 sections
59.41 GB

Volumetric genomic data

  Added
  Journal / Date
  Dataset
  Title
  Modality
  Resolution
  Organ
  Volumes
  Z-levels
  Num. spots/cells
  Size




  07/15/26
  Nature Methods (November 2025)

    Deep-STARmap

    [Paper]

  Scalable spatial single-cell transcriptomics and translatomics in 3D thick tissue blocks
  3D Spatial Transcriptomics (Deep-STARmap) and proteomics (Deep-RIBOmap)
  Single-cell 3D ST/SP data 
  Mouse brain and human cutaneous squamous cell carcinoma (cSCC) tissue
  2
  cSCC: 60µm depth, mouse brain: 150µm depth 
  cSCC: N/A cells,254 genes; Mouse brain: 198,675 cells, 1,017 genes
  2.1GB


  07/15/26
  Nature Methods (September 2025)

    3Dniches

    [Paper]

  Highly multiplexed 3D profiling of cell states and immune niches in human tumors
  3D CyCIF
  140 nm × 140 nm × 280 nm voxels (200–500 voxels per cell) - 20–54-plex 
  Melanoma specimens: (1) preinvasive cutaneous melanoma, (2) invasive VGP primary melanoma, (3) Metastatic melanoma to the skin
  3
  5–50-µm-thick sections 
  N/A
  3.17TB


  07/15/26
  Nature Methods (December 2025)

    SpatialZ

    [Paper]

  Bridging the dimensional gap from planar spatial transcriptomics to 3D cell atlases
  3D Spatial Transcriptomics with StarMAP
  Single-cell 3D ST data 
  Mouse visual cortex
  1
  89
  32,845 cells
  30.5MB


  07/15/26
  Nature Metabolism (March 2025)

    MetaVision3D

    [Paper]

  AI-driven framework to map the brain metabolome in three dimensions
  3D metabolomics (MALDI)
  50 µm 
  Brain mice samples: two wild-type, 5xFAD, and GAA 
  4
  79 serial brain sections (10-µm-thick sections). Sections are 50 µm apart (z axis), corresponding to the MALDI imaging spatial resolution of 50 µm (x and y axes).  
  N/A
  3.3GB


  07/15/26
  Nature Cancer (January 2022)

    3D IMC

    [Paper]

  Three-dimensional imaging mass cytometry for highly multiplexed molecular and cellular mapping of tissues and the tumor microenvironment
  3D proteomics (IMC)
  Single-cell resolution (40 markers) 
  HER2+ ductal breast carcinoma sample
  4 (Size of the main sample: 652×488×304µm3)
  152 consecutive slices (2-µm-thick sections)
  N/A
  6.6GB

Volumetric morphological and genomic data

  Added
  Journal / Date
  Dataset
  Title
  Modality
  Resolution
  Organ
  Volumes
  Serial sections
  Num. spots/cells
  Size




  08/25/26
  Cell systems (May 2025)

    3D NSCLC CosMx + ECM/SHG

    [Paper]

  Combining spatial transcriptomics and ECM imaging in 3D for mapping cellular interactions in the tumor microenvironment
  Serial sections; CosMx spatial transcriptomics, SHG, H&E and IF
  5 µm section thickness; CosMx sections spaced 30 µm; single-cell ST
  Human NSCLC lung tumor
  1 patient / 1 tumor
  34 total;

6 profiled with CosMx
340,644 cells
9GB

  07/15/26
  Nature (October 2024)

    2D3DTumorEvolution

    [Paper]

  Tumour evolution and microenvironment interactions in 2D and 3D space
  Serial H&E sections and Visium ST
  20x,55μm-diameter ST spots
  BRCA,CHOL,CRC,PDAC,RCC,UCEC
  78 patients (combined 2D and 3D) - Detailed description 
  131
  433,000 spots 
  N/A


  07/15/26
  Cell (July 2024)

    OpenST

    [Paper]

  Open-ST: High-resolution spatial transcriptomics in 3D
  Serial H&E sections and OpenST transcriptomics
  0.345 μm/pixel, single-cell ST data
  Head-and-neck metastatic lymph node
  1 (1 patient)
  19
  1,097,769 cells
  2.2GB


  07/15/26
  Nature Communications (February 2022)

    3DST_RA

    [Paper]

  Three-dimensional spatial transcriptomics uncovers cell type localizations in the human rheumatoid arthritis synovium
  Serial H&E sections and Spatial Transcriptomics
  20x, Spatial Transcriptomics technology (~100μm-diameter spots)
  Synovial joints (Rheumatoid arthritis)
  6 (6 patients)
  27 (RA1: 4, RA2: 7, RA3: 4, RA4: 3, RA5: 5, RA6: 4). RA1 spacing between sections -> 21μm ; RA2~RA6: consecutive sections  
  17,117 spots
  2.2GB (.gz format)


  07/15/26
  Nature Neuroscience (March 2021)

    Human prefrontal cortex

    [Paper]

  Transcriptome-scale spatial gene expression in the human dorsolateral prefrontal cortex
  Serial H&E sections and Visium ST
  Visium ST technology (55μm-diameter spots)
  Human dorsolateral prefrontal cortex
  3 patient
  12 (4 sections for each patient; The distance between sections is 10μm - 300μm - 10μm)
  47,681 spots
  N/A


  07/15/26
  Nature Communications (February 2021)

    HER2+

    [Paper]

  Spatial deconvolution of HER2-positive breast cancer delineates tumor-associated cell type interactions
  Serial H&E sections and Spatial Transcriptomics
  20x, Spatial Transcriptomics technology (~100μm-diameter spots)
  HER2+ breast cancer
  8 (8 patients)
  36 (A~D:6, E~H:3). Spacing between sections: 36μm. Section thickness: 16μm
  13,299 spots
  629.6 MB


  07/15/26
  Cell (December 2019)

    Human Heart

    [Paper]

  A Spatiotemporal Organ-Wide Gene Expression and Cell Atlas of the Developing Human Heart
  Serial H&E sections and Spatial Transcriptomics
  20x, Spatial Transcriptomics technology (~100μm-diameter spots)
  Developing Human Heart
  Four developmental heart tissues
  19 (4,9,6 sections for 4.5–5, 6.5, and 9 PCW heart tissues). Spacing between sections: 36μm. Section thickness: 5μm
   3,115 spots
  211 MB

Contributions

If you would like to add a 3D dataset to the current compilation, please create a pull request with details for each field in the tables above, or contact [email protected] .