{"success":true,"data":{"author":"Eva Blondeel, Filippo Piccinini, Olivier De Wever et al.; mirrored by MedOtter","bucketAccessType":"public","bucketName":"MedOtter/SLiMIA","bucketPrefix":"","category":"Life Sciences","cloudProvider":"huggingface","createdAt":"2026-11-21T20:45:56.195Z","description":"SLiMIA (Spheroid Light Microscopy Image Atlas) is a multi-lab collection of transmitted-light images of three-dimensional cell culture spheroids, tumor and non-tumor, made for morphometric analysis such as spheroid size, area and shape. This mirror holds 7,984 image and manual mask pairs acquired on nine microscopes (five phase-contrast and four brightfield) and covering 48 cell lines, 8 culture media and 4 spheroid formation methods. Masks are binary (0 or 255) with a single spheroid foreground class; image sizes range from 800x600 to 2592x1944 pixels. There is no upstream train/test split, so all pairs are in train, and the microscope and originating lab are recorded per row, which matters because they introduce batch effects. Twelve unpaired upstream files are listed in unpaired.json and scale-bar images are in scalebars/. Parquet under data/, about 8.4 GB (measured 2026-11-21); read with Hugging Face datasets or pandas plus Pillow. Use for spheroid segmentation, organoid and tumor spheroid size measurement and cross-microscope generalization. The nine figshare data articles of collection DOI 10.6084/m9.figshare.c.7486311.v1 are CC BY 4.0 (the companion paper itself is CC BY-NC-ND); cite Blondeel et al., Scientific Data 2025 (doi 10.1038/s41597-025-04441-x).","displayTitle":"SLiMIA: Spheroid Light Microscopy Image Atlas, 7,984 3D cell culture spheroid images with masks from 9 microscopes","doi":"10.6084/m9.figshare.c.7486311.v1","huggingFaceUrl":"https://huggingface.co/datasets/MedOtter/SLiMIA","id":"e73c31db-664e-5eb9-bbee-c57a97aeea95","license":"CC-BY-4.0","licenseEvidenceUrl":"https://figshare.com/articles/dataset/27390402","objectCountEstimate":59,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1038/s41597-025-04441-x","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/SLiMIA\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"MedOtter/SLiMIA\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=MedOtter, license:cc-by-4.0, size_categories:1K<n<10K, format:parquet, downloads=133"],"format":"mixed","generatedAt":"2026-11-21T20:48:39.823Z","layout":{"example":"data/train-00000-of-00042.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"data/train-00000-of-00042.parquet","truncatedAt":512}],"sampleObjects":["data/train-00000-of-00042.parquet","data/train-00001-of-00042.parquet","data/train-00002-of-00042.parquet","data/train-00003-of-00042.parquet","data/train-00004-of-00042.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n- image-segmentation\ntags:\n- medical-imaging\n- microscopy\n- spheroid\n- 3d-cell-culture\n- brightfield\n- phase-contrast\nsize_categories:\n- 1K<n<10K\n---\n\n# SLiMIA - Spheroid Light Microscopy Image Atlas\n\nBinary segmentation of 3D tumour and non-tumour **spheroids** in transmitted-light\nmicroscopy, spanning **9 microscopes, 48 cell lines, 8 culture media and 4 spheroid\nformation methods**. Mirror of figshare collection\n[7486311](https://figshare.com/collections/The_Spheroid_Light_Microscopy_Image_Atlas_SLiMIA_for_morphometrical_analysis_of_three_dimensional_cell_cultures/7486311).\n\n## License\n\n**CC BY 4.0** - all nine figshare data articles carry it (verified against the\nfigshare API `license` field per article and against DataCite SPDX).\n\n> Note the trap: the accompanying *Scientific Data* **paper** is CC BY-NC-ND 4.0.\n> That covers the article text and figures, **not** the data. Several public\n> dataset indexes copied the article licence onto the dataset and wrongly list\n> SLiMIA as non-redistributable.\n\n## Contents\n\n| | |\n|---|---|\n| Pairs | **7,984** (image + manual mask) |\n| Modality | transmitted light: phase-contrast (5 scopes) / brightfield (4 scopes) |\n| Target | single foreground class - spheroid |\n| Mask values | `{0, 255}` binary |\n| Resolution | 800x600 to 2592x1944 |\n| Split | none upstream - everything is `train` |\n\n### Per-microscope\n\n| Microscope | Contrast | Pairs |\n|---|---|---|\n| Axiovert200 | phase-contrast | 704 |\n| Axiovert200M | phase-contrast | 3,107 |\n| Cytation5 | brightfield | 66 |\n| IncucyteS3 | phase-contrast | 468 |\n| IncucyteZOOM | phase-contrast | 742 |\n| LeicaDMi1 | brightfield | 2,311 |\n| LeicaDMi3000B | phase-contrast | 439 |\n| OlympusIX05 | brightfield | 88 |\n| OperaPhenix | brightfield | 59 |\n\n`Axiovert200M` is additionally split by originating lab in the `lab` column\n(DeWever 2,558 / Kunz-Schughart 549) - a real batch effect worth stratifying on.\n\n## Why 7,984 and not 7,990\n\nThe atlas nominally holds 7,990 images and 7,990 masks. Six LeicaDMi1 images have\nno mask and six LeicaDMi1 masks have no image - genuinely different `B*`/`T*`\ncombinations, not a name-matching failure. An image without ground truth is not\nevaluable, so all 12 are omitted and enumerated in `unpaired.json`.\n\nFull accounting of the 15,992 upstream files:\n`7,984x2 paired + 6 + 6 unpaired + 11 scalebars + 1 readme = 15,992`.\n\n## Ground truth\n\nA **single tier**: manual foreground/background annotation in **AnaSP v1.4**,\ntraced on enlarged images. There is no automated tier, no expert-corrected second\npass and no multi-annotator subset, so **no inter-rater agreement can be\ncomputed**. Annotation was distributed across the authors plus seven named\nhelpers, but per-image annotator identity is not recorded - annotator effects are\nunmodelable. The only documented QC is a two-observer *image-quality* screen\n(focus, borders, dust), not a mask review.\n\n## Empty masks are a real class, not missing annotation\n\n**107 of the 7,984 masks are entirely zero, and every one of them is IncucyteS3**\n(107 of that microscope's 468 frames, 22.9%). This is not corruption and not\nun-annotated data - it is spheroid formation kinetics, measured:\n\n* all 107 fall at **t = 0-18 h**; every non-empty mask begins at **14 h or later**\n* 107 of the 125 IncucyteS3 frames at t <= 18 h are empty (85.6%)\n* all are the same condition - **MCF10A seeded at 3,000 cells**\n* the transition is gradual across replicates (14 h: 1 non-empty, 15 h: 2,\n  16 h: 4, 17 h: 5, 18 h: 6) - i.e. the spheroid progressively aggregates\n\nSo an empty mask means **\"no spheroid has formed yet\" - a confirmed true\nnegative**. Treat it as a genuine negative, the inverse of datasets where an\nabsent mask means \"this class was not annotated\".\n\n**Evaluation consequence.** MedOtter's harness *skips* empty-reference frames\nrather than scoring them 0 - measured on this mirror: `num_images 7984,\nnum_processed 7877, num_skipped 107`, exactly the empty set - so they do not\ncorrupt the reported metric. They are equally **not evaluated**: the IncucyteS3\nnumber","encoding":"utf8","key":"README.md","name":"README.md","size":7921,"truncated":true}],"sourceObjectKey":"data/train-00000-of-00042.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/SLiMIA\")\nprint(ds)","starterSamplePath":"data/train-00000-of-00042.parquet","totalObjects":59},"summary":"SLiMIA (Spheroid Light Microscopy Image Atlas) is a multi-lab collection of transmitted-light images of three-dimensional cell culture spheroids, tumor and non-tumor, made for…","title":"SLiMIA: Spheroid Light Microscopy Image Atlas, 7,984 3D cell culture spheroid images with masks from 9 microscopes","updatedAt":"2026-11-21T21:58:09.060Z","visibility":"public","websiteUrl":"https://doi.org/10.6084/m9.figshare.c.7486311.v1"}}