OSCR

3D Ultrastructural segmentation of human microvasculature reveals new insights in pericyte-endothelial cell interactions.

Code ↔ Paper

1 match between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 1 match
  1. [1] § Methods › Using the 3D-Unet for semantic segmentation ↔ DeepLearningTrainingMatlab/main_train3dUnet_7class.m, lines 58–69 · score 0.81 · max epochs, mini batch, validation frequency, trained, Unet

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

MATLAB · 74 lines · 2.7 KB · no license · 1 match

  1. % get directory name where the training data is found
  2. imageDir = fullfile('C:\Users\TEV4\DLtraining\','VesTS_3Dmulticlass');
  4. %% Create data stores for Training data
  5. % create an image datastore
  6. volReader = @(x) matRead(x);
  7. volLoc = fullfile(imageDir,'imagesTr');
  8. volds = imageDatastore(volLoc, ...
  9. 'FileExtensions','.mat','ReadFcn',volReader);
  11. % create a pixelLabel datastore for Training data
  12. lblLoc = fullfile(imageDir,'labelsTr');
  13. classNames = ["background","BM","Lumen","Nuclei","Mito","ER","Unknown"];
  14. pixelLabelID = [0 1 2 3 4 5 6];
  15. pxds = pixelLabelDatastore(lblLoc,classNames,pixelLabelID, ...
  16. 'FileExtensions','.mat','ReadFcn',volReader);
  18. % create random patch datastore for Training data
  19. patchSize = [164 164 100];
  20. patchPerImage = 60;
  21. miniBatchSize = 12;
  22. patchds = randomPatchExtractionDatastore(volds,pxds,patchSize, ...
  23. 'PatchesPerImage',patchPerImage);
  24. patchds.MiniBatchSize = miniBatchSize;
  26. %% Create datastores for Validation data
  27. % repeat creating datastores for the validation data
  28. volLocVal = fullfile(imageDir,'imagesVal');
  29. voldsVal = imageDatastore(volLocVal, ...
  30. 'FileExtensions','.mat','ReadFcn',volReader);
  32. lblLocVal = fullfile(imageDir,'labelsVal');
  33. pxdsVal = pixelLabelDatastore(lblLocVal,classNames,pixelLabelID, ...
  34. 'FileExtensions','.mat','ReadFcn',volReader);
  36. dsVal = randomPatchExtractionDatastore(voldsVal,pxdsVal,patchSize, ...
  37. 'PatchesPerImage',patchPerImage);
  38. dsVal.MiniBatchSize = miniBatchSize;
  40. %% Create Unet Layers
  41. numChannels = 1;
  42. inputPatchSize = [patchSize numChannels];
  43. numClasses = 7;
  44. [lgraph,outPatchSize] = unet3dLayers(inputPatchSize,numClasses,'ConvolutionPadding','valid');
  46. %% data augmentation
  47. dataSource = 'Training';
  48. dsTrain = transform(patchds,@(patchIn)augmentAndCrop3dPatch(patchIn,outPatchSize,dataSource));
  50. dataSource = 'Validation';
  51. dsVal = transform(dsVal,@(patchIn)augmentAndCrop3dPatch(patchIn,outPatchSize,dataSource));
  54. inputLayer = image3dInputLayer(inputPatchSize,'Normalization','none','Name','ImageInputLayer');
  55. lgraph = replaceLayer(lgraph,'ImageInputLayer',inputLayer);
  58. %% training options
  59. options = trainingOptions('adam', ...
  60. 'MaxEpochs',200, ...
  61. 'InitialLearnRate',5e-4, ...
  62. 'LearnRateSchedule','piecewise', ...
  63. 'LearnRateDropPeriod',5, ...
  64. 'LearnRateDropFactor',0.95, ...
  65. 'ValidationData',dsVal, ...
  66. 'ValidationFrequency',400, ...
  67. 'Plots','training-progress', ...
  68. 'Verbose',false, ...
  69. 'MiniBatchSize',miniBatchSize);
  71. %% train network
  72. modelDateTime = string(datetime('now','Format',"yyyy-MM-dd-HH-mm-ss"));
  73. [net,info] = trainNetwork(dsTrain,lgraph,options);
  74. save(strcat("trained3DUNet-",modelDateTime,"-Epoch-",num2str(options.MaxEpochs),".mat"),'net');

main_train3dUnet_7class.m at commit f873001, no license · at the source

Overview

Authors: Timothy E. Vanderleest1, Joseph F. Arboleda-Velasquez1
  1. Schepens Eye Research Institute of Mass Eye and Ear and the Department of Ophthalmology at Harvard Medical School, Boston, MA, United States
Institutions: Massachusetts Eye and Ear Infirmary (United States); Harvard University (United States)
Journal: Frontiers in cell and developmental biology, volume 14, article 1788958
Dates: received 15 January 2026; accepted 15 April 2026; published online 12 May 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.3389/fcell.2026.1788958 · PMID 42205677 · PMCID PMC13201225 · OpenAlex W7160920388
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: human (organism), cellular / molecular (subfield)
Methods: Statistics
Keywords: 3D reconstruction, 3D U-net, endothelial cells, pericytes, ultrastructure
Topic: Barrier Structure and Function Studies (Neurology, Neuroscience), according to OpenAlex
Funding: Good Ventures Foundation
Citations: not cited yet (Europe PMC); 25 references in the paper

Abstract

Introduction: Capillary structure remains incompletely characterized at the nanoscale level. Advances in 3D electron microscopy datasets provide new opportunities to systematically examine microvascular architecture and cellular interactions.

Methods: We performed 3D image segmentation of human cerebral cortex microvasculature using the publicly available H01 Release dataset. This initial analysis focused on descriptive and quantitative characterization of the basement membrane, peg-and-socket cell interactions, and other subcellular features.

Results: We identified several novel structural features. In addition to known bidirectional peg-and-socket connections between pericytes (PCs) and endothelial cells (ECs), we observed similar structures between neighboring ECs and at interfaces where ECs form junctions with themselves. PC pegs showed an unexpected preference for proximity to EC nuclei and were enriched at PC edges. The PC endoplasmic reticulum (ER) frequently contacted the plasma membrane on the lumen-facing surface, particularly at sites corresponding to sockets of EC pegs. We also identified electron-lucent pockets (ELPs) at interfaces between ECs and PCs, as well as within the basement membrane.

Discussion: These findings expand current understanding of capillary ultrastructure by revealing previously unrecognized interactions between vascular cells and novel subcellular features. The observed structural features may have functional implications for signaling, stability, and vascular mechanics, warranting further investigation.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 1 match between paragraphs and lines of code.

timvanderleest/Vessel-Ultrastructure-Analysis

License: none: the authors keep all their rights
State: the link answers, verified on 28 September 2026
Evidence: files inventoried
Commit: f873001e5e1f87622b7f0996a7c58ba0eda556c4, 5 April 2026
Languages: MATLAB (3), Python (1)
Size: 6 files, 4 scripts
Software Heritage: not archived
Found in: “Data availability statement”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Deep Learning Toolbox (1 file), NumPy (1 file), tifffile (1 file)
Availability: 1 check, the latest on 28 September 2026: the link answers
  • 28 September 2026: the link answers
5 files
At the source: github.com/timvanderleest/Vessel-Ultrastructure-Analysis

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 4 scripts, each with its path and the digest of its content;
  • 1 match between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

Matlab codes developed for our analysis pipeline are available through Github (https://github.com/timvanderleest/Vessel-Ultrastructure-Analysis). The annotated image data developed in this study have been made available through the Bioimage Archive (https://www.ebi.ac.uk/bioimage-archive) under accession number S-BIAD3187 (Sarkans et al., 2018). The associated raw image data can be obtained by downloading from Google Cloud Storage using a Python script in the Github repository. The trained 3D U-Net model, which was trained using Matlab Deep Learning Toolbox, has been uploaded to the Matlab File Exchange (https://www.mathworks.com/matlabcentral/fileexchange/183578-trained-3d-u-net-for-a-7-class-microvessel-segmentation). An image showing the 3D U-Net training progress metrics (Accuracy and Loss) is also available in the Github repository.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 2, 28 September 2026

  • Funding: added Good Ventures Foundation

Version 1, 28 September 2026: the first record

Recorded: type, language, journal, volume, pages, dates, 2 authors, 5 keywords, 25 references.

Cite

This paper

Vanderleest, T. E., & Arboleda-Velasquez, J. F. (2026). 3D Ultrastructural segmentation of human microvasculature reveals new insights in pericyte-endothelial cell interactions. Frontiers in cell and developmental biology, 14, 1788958. https://doi.org/10.3389/fcell.2026.1788958

BibTeX

@article{vanderleest20263d,
author = {Vanderleest, Timothy E. and Arboleda-Velasquez, Joseph F.},
title = {{3D Ultrastructural segmentation of human microvasculature reveals new insights in pericyte-endothelial cell interactions}},
journal = {Frontiers in cell and developmental biology},
year = {2026},
month = may,
volume = {14},
pages = {1788958},
publisher = {Frontiers Media SA},
issn = {2296-634X},
doi = {10.3389/fcell.2026.1788958},
url = {https://doi.org/10.3389/fcell.2026.1788958},
pmid = {42205677},
pmcid = {PMC13201225}
}

RIS

TY - JOUR
AU - Vanderleest, Timothy E.
AU - Arboleda-Velasquez, Joseph F.
TI - 3D Ultrastructural segmentation of human microvasculature reveals new insights in pericyte-endothelial cell interactions
T2 - Frontiers in cell and developmental biology
J2 - Front Cell Dev Biol
PY - 2026
DA - 2026/05/12
VL - 14
SP - 1788958
SN - 2296-634X
PB - Frontiers Media SA
DO - 10.3389/fcell.2026.1788958
UR - https://doi.org/10.3389/fcell.2026.1788958
LA - en
ER -

CSL-JSON

{
"id": "10.3389/fcell.2026.1788958",
"type": "article-journal",
"title": "3D Ultrastructural segmentation of human microvasculature reveals new insights in pericyte-endothelial cell interactions",
"container-title": "Frontiers in cell and developmental biology",
"author": [
{
"family": "Vanderleest",
"given": "Timothy E."
},
{
"family": "Arboleda-Velasquez",
"given": "Joseph F."
}
],
"container-title-short": "Front Cell Dev Biol",
"volume": "14",
"page": "1788958",
"DOI": "10.3389/fcell.2026.1788958",
"PMID": "42205677",
"PMCID": "PMC13201225",
"ISSN": "2296-634X",
"publisher": "Frontiers Media SA",
"URL": "https://doi.org/10.3389/fcell.2026.1788958",
"language": "en",
"issued": {
"date-parts": [
[
2026,
5,
12
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s41467-026-71458-0 [code]
Early differential impact of MeCP2 mutations on functional networks in Rett syndrome patient-derived human cortical organoids.
Journal: Nature communications
In common: Deep Learning Toolbox, tifffile, NumPy, cellular / molecular
[2] doi:10.1038/s41593-026-02388-9 [code]
Hippocampal CA3 connectomics reveals a gradient of mossy fiber inputs and selective feedforward inhibition onto pyramidal cells.
Journal: Nature neuroscience
In common: NumPy, 3 references
[3] doi:10.1126/sciadv.aec4911 [code]
Dendritic shaft constrictions shape synaptic integration in neurons.
Journal: Science advances
In common: NumPy, cellular / molecular, 2 references
[4] doi:10.1073/pnas.2533168123 [code]
Dendritic morphology and synaptic nonlinearities enhance functional complexity in human cortical neurons.
Journal: Proceedings of the National Academy of Sciences of the United States of America
In common: NumPy, cellular / molecular, 2 references
[5] doi:10.1073/pnas.2523612123
A pericyte chloride clamp mechanism governs capillary control of cerebral blood flow.
Journal: Proceedings of the National Academy of Sciences of the United States of America
In common: cellular / molecular, 2 references
[6] doi:10.21037/qims-2026-0792 [code]
An nnU-Net-based framework with adaptive feature representation for 3D brain tumor segmentation.
Journal: Quantitative imaging in medicine and surgery
In common: tifffile, NumPy, 1 reference
[7] doi:10.1364/boe.600665 [code]
NeuroSeg-MF: robust neuron segmentation in two-photon Ca<sup>2+</sup> imaging using multi-feature fusion and detection-guided SAM.
Journal: Biomedical optics express
In common: tifffile, NumPy, 1 reference
[8] doi:10.1002/hbm.70553 [code]
Axon Diameter Mapping in the Living Human Brain with Ultra-High-Gradient Diffusion MRI at 500 mT/m Gradient Strength.
Journal: Human brain mapping
In common: Deep Learning Toolbox, NumPy
[9] doi:10.1038/s41593-026-02258-4 [code]
Laminar organization of cellular microcircuits modulating human interictal epileptiform discharges.
Journal: Nature neuroscience
In common: Deep Learning Toolbox, NumPy
[10] doi:10.1371/journal.pcbi.1013499 [code]
VesiclePy: A machine learning vesicle analysis toolbox for volume electron microscopy.
Journal: PLoS computational biology
In common: NumPy, 2 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.