{"success":true,"data":{"author":"Emna Harigua, Rafeh Oualha, Nesrine Fekih-Romdhane et al.; curated by SPORA","bucketAccessType":"public","bucketName":"csmp-hub/air-leish","bucketPrefix":"","category":"Health & Biomedical","cloudProvider":"huggingface","createdAt":"2026-11-17T19:29:33.239Z","description":"AIR-LEISH is an AI-ready microscopy dataset for counting and segmenting intracellular Leishmania parasites. It contains 180 Giemsa-stained bright-field images of infected macrophages with expert-drawn outlines of nuclei, whole cells and the small parasites inside them: 90 images of THP-1-derived macrophages infected with Leishmania major (Set1) and 90 of primary human monocyte-derived macrophages infected with Leishmania infantum (Set2). These are laboratory cell preparations, not tissue sections. train.parquet holds all 180 images with masks, and Set1.coco.json and Set2.coco.json keep the polygon annotations in COCO format. The released files do not identify the paper's random train/validation/test assignment, and some images show the same cells in shifted views, so random image splits can leak; shared-view-review.json documents these. About 890 MB (measured 2026-11-17). Images and masks are stored as PNG or TIFF bytes inside Parquet; read them with polars or pandas plus Pillow, or with Hugging Face datasets. Use for parasite detection and counting, infection rate estimation and instance segmentation. Curated and packaged on Hugging Face by the SPORA project (csmp-hub), which keeps the original license and citation files. The authors' Zenodo record 17384855 is CC BY 4.0; cite Harigua, Oualha, Fekih-Romdhane et al., Scientific Data 2026 (doi 10.1038/s41597-026-06676-8).","displayTitle":"AIR-LEISH: Giemsa-stained microscopy of Leishmania-infected macrophages with nucleus, cell and parasite outlines","doi":"10.5281/zenodo.17384855","huggingFaceUrl":"https://huggingface.co/datasets/csmp-hub/air-leish","id":"b004765d-c103-5f04-b8ba-e3f711dc4e14","license":"CC-BY-4.0","licenseEvidenceUrl":"https://zenodo.org/records/17384855","objectCountEstimate":13,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1038/s41597-026-06676-8","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"csmp-hub/air-leish\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"csmp-hub/air-leish\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=csmp-hub, license:cc-by-4.0, size_categories:n<1K, format:parquet, downloads=196"],"format":"mixed","generatedAt":"2026-11-17T19:40:22.592Z","layout":{"example":"train.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"train.parquet","truncatedAt":512}],"sampleObjects":["NOTICE.md","Set1.coco.json","Set2.coco.json","examples.png","masks.png"],"sidecars":[{"bytes":"Attribution 4.0 International\n\n=======================================================================\n\nCreative Commons Corporation (\"Creative Commons\") is not a law firm and\ndoes not provide legal services or legal advice. Distribution of\nCreative Commons public licenses does not create a lawyer-client or\nother relationship. Creative Commons makes its licenses and related\ninformation available on an \"as-is\" basis. Creative Commons gives no\nwarranties regarding its licenses, any material licensed under their\nterms and conditions, or any related information. Creative Commons\ndisclaims all liability for damages resulting from their use to the\nfullest extent possible.\n\nUsing Creative Commons Public Licenses\n\nCreative Commons public licenses provide a standard set of terms and\nconditions that creators and other rights holders may use to share\noriginal works of authorship and other material subject to copyright\nand certain other rights specified in the public license below. The\nfollowing considerations are for informational purposes only, are not\nexhaustive, and do not form part of our licenses.\n\n     Considerations for licensors: Our public licenses are\n     intended for use by those authorized to give the public\n     permission to use material in ways otherwise restricted by\n     copyright and certain other rights. Our licenses are\n     irrevocable. Licensors should read and understand the terms\n     and conditions of the license they choose before applying it.\n     Licensors should also secure all rights necessary before\n     applying our licenses so that the public can reuse the\n     material as expected. Licensors should clearly mark any\n     material not subject to the license. This includes other CC-\n     licensed material, or material used under an exception or\n     limitation to copyright. More considerations for licensors:\n    wiki.creativecommons.org/Considerations_for_licensors\n\n     Considerations for the public: By using one of our public\n     licenses, a licensor grants the public permission to use the\n     licensed material under specified terms and conditions. If\n     the licensor's permission is not necessary for any reason--for\n     example, because of any applicable exception or limitation to\n     copyright--then that use is not regulated by the license. Our\n     licenses grant only permissions under copyright and certain\n     other rights that a licensor has authority to grant. Use of\n     the licensed material may still be restricted for other\n     reasons, including because others have copyright or other\n     rights in the material. A licensor may make special requests,\n     such as asking that all changes be marked or described.\n     Although not required by our licenses, you are encouraged to\n     respect those requests where reasonable. More considerations\n     for the public:\n    wiki.creativecommons.org/Considerations_for_licensees\n\n=======================================================================\n\nCreative Commons Attribution 4.0 International Public License\n\nBy exercising the Licensed Rights (defined below), You accept and agree\nto be bound by the terms and conditions of this Creative Commons\nAttribution 4.0 International Public License (\"Public License\"). To the\nextent this Public License may be interpreted as a contract, You are\ngranted the Licensed Rights in consideration of Your acceptance of\nthese terms and conditions, and the Licensor grants You such rights in\nconsideration of benefits the Licensor receives from making the\nLicensed Material available under these terms and conditions.\n\n\nSection 1 -- Definitions.\n\n  a. Adapted Material means material subject to Copyright and Similar\n     Rights that is derived from or based upon the Licensed Material\n     and in which the Licensed Material is translated, altered,\n     arranged, transformed, or otherwise modified in a manner requiring\n     permission under the Copyright and Similar Rights held by the\n     Licensor. For purposes of this Public License, where the Licensed\n     Material is a musical work, performance, or sound recording,\n ","encoding":"utf8","key":"LICENSE","name":"LICENSE","size":18657,"truncated":true},{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n- image-segmentation\ntags:\n- microscopy\n- brightfield\n- nuclei\n- whole-cell\n- parasites\n- instance-segmentation\nsize_categories:\n- n<1K\nconfigs:\n- config_name: default\n  data_files:\n  - split: train\n    path: train.parquet\n---\n\n# AIR-LEISH\n\n## Overview\n\nAIR-LEISH contains **180 Giemsa-stained microscope images of macrophages infected with Leishmania**, with expert-drawn outlines for nuclei, whole cells and the small parasites inside them. It can support cell segmentation, parasite counting and studies of infection.\n\nThere are 90 images of THP-1-derived macrophages infected with *L. major* (`Set1`), and 90 images of primary human monocyte-derived macrophages infected with *L. infantum* (`Set2`). These are laboratory cell preparations, rather than tissue sections.\n\n**License: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).** You may use, change and share the data, including commercially, if you credit the creators, keep the license notices and describe your changes.\n\n## Files and splits\n\n| Split | File | Images |\n| --- | --- | ---: |\n| `train` | `train.parquet` | 180 |\n\nAll images are kept together. The paper describes random training, validation and test sets, but the released files do not identify those assignments or which images share a slide or donor. **There is no independent validation or test split in this version.** Some images show the same cells in shifted views, confirmed in both infection models. A random image split can therefore put related views into both training and evaluation.\n\n`Set1.coco.json` and `Set2.coco.json` preserve the original outlines and source metadata. License and annotation/split notes are included alongside the Parquet file.\n\n## Fields\n\nEach row contains one image and all its annotations.\n\n| Field | Contents |\n| --- | --- |\n| `id` | Unique image name, such as `air_leish_set1_000` |\n| `image` | Colour image stored as PNG bytes; HWC RGB uint8, values 0–255 |\n| `inst` | Numbered nucleus mask, stored as uint16 PNG bytes |\n| `inst_cell` | Numbered whole-cell mask, stored as uint16 PNG bytes |\n| `inst_parasite` | Numbered parasite mask, stored as uint16 PNG bytes |\n| `type` | Nucleus category mask: 3 inside a nucleus, 0 elsewhere; uint8 PNG bytes |\n| `pairs` | Suggested nucleus-to-cell matches; several nuclei may match the same cell |\n| `objects` | Every individual object mask, including overlapping pixels; details below |\n| `source_mask` | Original combined annotation PNG bytes; `null` for the one image without this file |\n| `source_set`, `infection_model` | Original collection and its cell/parasite model |\n| `file_name`, `source_file_name` | Actual archive filename and original COCO filename |\n| `source_image_id` | Original image number within `source_set` |\n\nImages are 2,709 pixels high × 1,844 wide; all masks use the same HW shape. Zero is background in the numbered masks. **Instance number = original annotation number + 1**, because the source uses annotation number 0. A nucleus and its cell can have different numbers.\n\n`objects` is a list with `instance_id`, `source_annotation_id`, `category_id` and `rle_counts`. The last field holds a complete binary mask in compressed COCO run-length encoding, using the image's height and width. The loading example reads one such mask. Object numbers are scoped to an image and source set.\n\n## Labels\n\n| Structure | Original objects | Visible in numbered masks |\n| --- | ---: | ---: |\n| Nuclei | 1,634 | 1,633 |\n| Whole cells | 1,427 | 1,427 |\n| Parasites | 7,994 | 7,994 |\n\nEvery original object is kept in `objects`, even when another object overlaps it. The three numbered masks give each pixel one object number within its layer. Where objects in the same layer overlap, the smaller object takes the shared pixels; the larger object number wins equal-area ties. One nucleus is completely covered by another in the numbered mask; both full masks remain in `objects`. Use `objects` when you need the full overlapping masks. The original polygon coordinates are also kept in the COCO files.\n\nThe source categories ","encoding":"utf8","key":"README.md","name":"README.md","size":12426,"truncated":true},{"bytes":"{\n  \"dataset\": \"AIR-LEISH\",\n  \"source_record\": \"https://zenodo.org/records/17384855\",\n  \"source_doi\": \"10.5281/zenodo.17384855\",\n  \"license\": \"CC-BY-4.0\",\n  \"source_archive\": {\n    \"file\": \"AIR_LEISH_dataset_v1.zip\",\n    \"sha256\": \"36cfeeecac27cc84266c40c8676c957065f4ddca2dfe9b231e86ad84d7451d2e\",\n    \"md5\": \"467fc13e79fb0eefae5e83fbd46a48e2\",\n    \"bytes\": 997142744\n  },\n  \"software\": {\n    \"numpy\": \"2.5.4\",\n    \"pillow\": \"12.3.0\",\n    \"pyarrow\": \"23.0.1\",\n    \"pycocotools\": \"2.0.10\"\n  },\n  \"conversion\": \"Opaque alpha removed; RGB pixels unchanged. Original polygons filled with pycocotools, without geometry repair. Separate uint16 nucleus/cell/parasite rasters, smaller raster area wins within-layer overlap; larger instance ID wins ties. Instance ID = source annotation ID + 1 because source ID 0 is valid. Every full individual COCO-fill mask is also stored in objects as compressed RLE; original COCO documents and semantic PNGs retained. type = original NU category 3 on nucleus foreground, 0 elsewhere. No resizing or patching.\",\n  \"pairing_method\": \"Inferred from full individual object masks: nonempty nucleus fully contained in exactly one positively overlapping cell. Multiple nuclei may link to the same cell; unmatched/partial/ambiguous objects retained. No expert pairing IDs provided.\",\n  \"split_method\": \"One train split containing all 180 images. No original split lists or slide/donor identifiers are present; no independent evaluation split claimed. Duplicate/near-duplicate/shared-view audit supplied for future grouping.\",\n  \"rows\": [\n    {\n      \"id\": \"air_leish_set1_000\",\n      \"source_set\": \"Set1\",\n      \"source_image_id\": 0,\n      \"file_name\": \"20250328_CCimage88.png\",\n      \"objects\": {\n        \"inst_parasite\": {\n          \"source_objects\": 12,\n          \"visible_objects\": 12,\n          \"overlap_pixel_assignments\": 0,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [],\n          \"hidden_object_ids\": []\n        },\n        \"inst_cell\": {\n          \"source_objects\": 1,\n          \"visible_objects\": 1,\n          \"overlap_pixel_assignments\": 0,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [],\n          \"hidden_object_ids\": []\n        },\n        \"inst\": {\n          \"source_objects\": 1,\n          \"visible_objects\": 1,\n          \"overlap_pixel_assignments\": 0,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [],\n          \"hidden_object_ids\": []\n        }\n      },\n      \"pairs\": 1,\n      \"nuclei_no_cell_overlap\": [],\n      \"nuclei_multiple_cell_overlaps\": [],\n      \"cells_overlapped_by_multiple_nuclei\": [],\n      \"nuclei_without_pair\": [],\n      \"pixel_sha256\": {\n        \"image\": \"de011e08956d26133d4e35a691ec6c218fe138f7642a1fad6d956008d203b9ae\",\n        \"type\": \"2a4994838f591d35b4a8219726b0b92500224b42cf1131aa92b5bba929dbe413\",\n        \"inst_parasite\": \"7550afa5526433760183b9c45ecd2be9e34d41c5add05503886193137c3d1861\",\n        \"inst_cell\": \"9bcef33d0fcf81b3b54d56f4917ed8e161197cadc0e9f7486eb7c2e91ed373f2\",\n        \"inst\": \"ae7fdea8a06d3f5d19dc8d12497099f7863bb68fa19fdba3f603dffd97d9622a\"\n      }\n    },\n    {\n      \"id\": \"air_leish_set1_001\",\n      \"source_set\": \"Set1\",\n      \"source_image_id\": 1,\n      \"file_name\": \"20250328_CCimage2.png\",\n      \"objects\": {\n        \"inst_parasite\": {\n          \"source_objects\": 25,\n          \"visible_objects\": 25,\n          \"overlap_pixel_assignments\": 0,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [],\n          \"hidden_object_ids\": []\n        },\n        \"inst_cell\": {\n          \"source_objects\": 12,\n          \"visible_objects\": 12,\n          \"overlap_pixel_assignments\": 23,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [\n            21\n          ],\n          \"hidden_object_ids\": []\n        },\n        \"inst\": {\n          \"source_objects\": 17,\n          \"visible_objects\": 17,\n          \"overlap_pixel_assignments\": 0,\n          \"empty_raster_ids\": [],\n          \"occluded_object_ids\": [],\n          \"hidden_object_ids\": []\n        }\n      },\n      \"pairs\": 16,\n      \"nuclei_no_cell_overlap\":","encoding":"utf8","key":"manifest.json","name":"manifest.json","size":287722,"truncated":true}],"sourceObjectKey":"train.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"csmp-hub/air-leish\")\nprint(ds)","starterSamplePath":"examples.png","totalObjects":13},"summary":"AIR-LEISH is an AI-ready microscopy dataset for counting and segmenting intracellular Leishmania parasites.","title":"AIR-LEISH: Giemsa-stained microscopy of Leishmania-infected macrophages with nucleus, cell and parasite outlines","updatedAt":"2026-11-17T23:08:51.206Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/csmp-hub/air-leish"}}