HeLaCytoNuc: fluorescence microscopy dataset with segmentation masks for cell nuclei and cytoplasm

The preprocessed dataset includes 2,676 8-bit RGB images, each with a pixel resolution of 520 x 696 pixels. In these images, only two of the RGB channels are utilized: the red channel represents the cytoplasm, and the blue channel represents the nuclei. The dataset is divided into training, validation, and test subsets in a 70:20:10 ratio.

The entire dataset is accompanied by instance segmentation masks for nuclei and cytoplasm objects obtained through a specialised CellProfiler [2] software. Notably, the test subset was annotated manually by a specialist, ensuring high-quality annotations. The original raw images are of a higher resolution, 1040 x 1392 pixels, and have a bit depth of 16 bits, providing more detailed information for advanced analyses.

File Description: The file structure of the zip files is as follows: HeLaCytoNuc_{train/validation/test}.zip -> - images -> {filename}.tif - nuclei_masks -> {filename}.tif - cytoplasm_masks -> {filename}.tif HeLaCytoNuc_raw_images.zip -> {filename}.tif HeLaCytoNuc_test_cellprofiler_masks.zip -> - nuclei_masks -> {filename}.tif - cytoplasm_masks -> {filename}.tif References: 1. Rämö, Pauli, Anna Drewek, Cécile Arrieumerlou, Niko Beerenwinkel, Houchaima Ben-Tekaya, Bettina Cardel, Alain Casanova et al. "Simultaneous analysis of large-scale RNAi screens for pathogen entry." BMC genomics 15 (2014): 1-18. 2.

Carpenter, Anne E., Thouis R. Jones, Michael R. Lamprecht, Colin Clarke, In Han Kang, Ola Friman, David A. Guertin et al. "CellProfiler: image analysis software for identifying and quantifying cell phenotypes." Genome biology 7 (2006): 1-11.