MoNuSeg (Multi-Organ Nucleus Segmentation)

H&E-stained histopathology images (from TCGA WSIs at 40x magnification) with per-nucleus binary segmentation masks. MICCAI 2018 challenge.

Overview

• Modality: H&E histopathology (brightfield microscopy)
• Image size: 1000x1000 RGB
• Samples: 37 train + 14 test = 51
• Organs (8 classes in tissue): 0 Unknown, 1 Breast, 2 Kidney, 3 Liver, 4 Prostate, 5 Bladder, 6 Colon, 7 Stomach. Test also includes lung and brain (labelled as tissue=0 Unknown here where not in the 8-class list).
• Ground truth: single-annotator semantic binary mask (0 = tissue, 1 = nucleus), derived by OR-combining all per-nucleus instance polygons.

Columns

Column	Type	Notes
patient	string	TCGA patient ID (e.g. TCGA-38-6178-01Z-00-DX1)
tissue	ClassLabel(8)	Organ label
image	Image (RGB)	1000x1000 H&E tile
mask	Image (mode 1)	1000x1000 binary nuclei mask
num_nuclei	int32	Instance count used to build the mask

Derivation

Source: RationAI/MoNuSeg parquet mirror of the Grand Challenge 2018 data. The instances column of the source (a list of per-nucleus binary PIL masks) was merged by logical OR to produce a semantic mask column. No other preprocessing.

License

CC BY-NC-SA 4.0. Underlying WSIs come from TCGA (public NIH data).

Citations

• Kumar et al., "A Dataset and a Technique for Generalized Nuclear Segmentation for Computational Pathology," IEEE TMI 36(7):1550-1560, 2017.
• Kumar et al., "A Multi-organ Nucleus Segmentation Challenge," IEEE TMI,