{"id":37702437,"url":"https://github.com/masilab/nucleus_and_cell_classification_on_he","last_synced_at":"2026-01-16T13:01:34.856Z","repository":{"id":239229118,"uuid":"798932562","full_name":"MASILab/nucleus_and_cell_classification_on_he","owner":"MASILab","description":"Classify nuclei/cells on intestinal H\u0026E into 14 different classes","archived":false,"fork":false,"pushed_at":"2025-01-30T22:37:04.000Z","size":1858,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-01-30T23:26:20.406Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MASILab.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-05-10T19:37:12.000Z","updated_at":"2025-01-30T22:37:08.000Z","dependencies_parsed_at":"2024-05-10T21:22:22.063Z","dependency_job_id":"1d165f90-0dfa-4e6b-8be9-0ce235f70f66","html_url":"https://github.com/MASILab/nucleus_and_cell_classification_on_he","commit_stats":null,"previous_names":["masilab/nucleus_and_cell_classification_on_he"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/MASILab/nucleus_and_cell_classification_on_he","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MASILab%2Fnucleus_and_cell_classification_on_he","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MASILab%2Fnucleus_and_cell_classification_on_he/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MASILab%2Fnucleus_and_cell_classification_on_he/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MASILab%2Fnucleus_and_cell_classification_on_he/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MASILab","download_url":"https://codeload.github.com/MASILab/nucleus_and_cell_classification_on_he/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MASILab%2Fnucleus_and_cell_classification_on_he/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28478887,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-16T11:59:17.896Z","status":"ssl_error","status_checked_at":"2026-01-16T11:55:55.838Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-01-16T13:01:34.191Z","updated_at":"2026-01-16T13:01:34.852Z","avatar_url":"https://github.com/MASILab.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n# Nucleus Segmentation \u0026 Subclassification on H\u0026E\n\n\u003cimg width=\"654\" alt=\"image\" src=\"https://github.com/user-attachments/assets/32372096-37fd-48c0-81f8-91deb485d4e7\" /\u003e\n\n\nPublication: [Data-driven nucleus subclassification on colon hematoxylin and eosin using style-transferred digital pathology](https://www.spiedigitallibrary.org/journals/journal-of-medical-imaging/volume-11/issue-6/067501/Data-driven-nucleus-subclassification-on-colon-hematoxylin-and-eosin-using/10.1117/1.JMI.11.6.067501.full)\n\n\n# This repo enables nucleus/cell classification on intestinal H\u0026E into 14 classes\n\n![Description of Image](./cell_classification.png)\n\nThis image shows a zoomed in section of a whole slide image of virtual H\u0026E, where nuclei/cells have been classified into 14 classes.\n\n# Note: \n**Pretrained weights are currently not public. They are only available to members of the MASI lab.**\n\n# Brief Overview\nThis repo provides pretrained models (as well as code to train new models).\nThe pretrained models were trained on virtual H\u0026E to classify nuclei into 14 classes.\n\n**If you want to use this model on real H\u0026E:**\nIt is recommended for the H\u0026E staining to be style transferred to the virtual H\u0026E style.\nYou can do that following instructions here: https://github.com/MASILab/he_stain_translation_cyclegan\n\nThen the nuclei must be located. \nYou can do that following instructions here: https://github.com/MASILab/hovernets_on_vhe\n\n\n\n# Citations\nIf you use this repo, please cite\n- \"Data-driven Nucleus Subclassification on Colon H\u0026E using Style-transferred Digital Pathology\"\n\n# Pretrained weights\nThe weights for 5 folds of trained resnets for nucleus classification on virtual H\u0026E can be found here:\n\n**MASI Lab:** ```/nfs/masi/remedilw/paper_journal_nucleus_subclassification/nucleus_subclassification/weights```\n\n**Public:** Not currently available\n\n# Inference\nAn example jupyter notebook showing how a pretrained model can be run on a whole slide image (virtual H\u0026E) can be found in this repo\n```inference_on_wsi.ipynb```\n\nIf you are in MASI lab, the paths are setup so that this inference notebook will run with the example data.\n\nIf you are not in MASI lab, change paths accordingly.\n\nThe pretrained models expect to perform inference on virtual H\u0026E or H\u0026E with resolution: 0.5 microns per pixel (mpp)\n\n# Training\nTraining can be run using\n```python __crossval_resnet_20k-steps_batch_256.py```\n\nIf you are in MASI lab, data necessary has been moved here:\n```/nfs/masi/remedilw/paper_journal_nucleus_subclassification/nucleus_subclassification/training_data```\n\nThe csvs and associated data are in that folder, though you will need to make a copy of the csvs and update the parent paths to the images once you move the data locally to a machine.\n\nIf you are not in the MASI lab, then the training script can be examined, but basically it expects\n1) Whole Slide Images of H\u0026E or virtual H\u0026E\n2) A csv, where each row is a nucleus, with a centroid (row and column), and a class label (cell type)\n\nThe resnet approach is simple. Given the coordinates of a nucleus, the dataloader reads a small patch around the nucleus, from the whole slide image (on the fly).\nThen it learns to predict the center nucleus in the patch, given the class label.\n\n# Note\nPlease note that in the provided scripts, the training involved on-the-fly resampling of virtual H\u0026E patches in the dataloader.\nHowever, in the provided inference script, the paths point to pre-resampled data, and so there is no on-the-fly resampling in the inference dataloader.\nPlease carefully consider the resolution you choose to operate on and adapt the code accordingly.\n\n# Working repo\nIf you are in the MASI lab, you can access my working github repo here\n\nhttps://github.com/MASILab/gca_he/tree/master/segmentation/vhe\n\nThis repo includes all code fragments etc.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmasilab%2Fnucleus_and_cell_classification_on_he","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmasilab%2Fnucleus_and_cell_classification_on_he","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmasilab%2Fnucleus_and_cell_classification_on_he/lists"}