2024_Hollandi-2_Embryo_masks

ARTICLE (when using these files, please, cite the following article):

Réka Hollandi, David Bauer, Akos Diosdi, Bálint Schrettner, Timea Toth, Dominik Hirling, Gábor Hollandi, Maria Harmati, József Molnár, Peter Horvath, "When the pen is mightier than the sword: semi-automatic 2 and 3D image labelling". (2024)

DESCRIPTION OF THE FILES:

1. 2D: Consists of HeLa Kyoto cells imaged by confocal images with multiple fluorescent channels including DAPI-405 (C001), Alexa Fluor-488 (C002), and Alexa Fluor-633 (C003). Each fluorescent image is 2048 × 2048 pixel resolution with 0.103 µm pixel size. All images were annotated by 3 experts, who annotated both the cytoplasm and nuclei of the images. Each of the images was labeled twice: once using only manual labor (manual) and again using Minimal contour (minimal). Each folder includes the annotation (Labels.tiff), the time for each annotation (annotation_times.csv), and a video file that visualizes the annotation process. Images are included.
2. Embryo: 3D fluorescent image data of a mouse embryo [1]. All three experts annotated the data separately. Each annotator did the manual annotation twice (manual and manual_2), while also annotating with the Minimal contour (minimal), and with the combination of Minimal contour and interpolation (manual_minimal_interpolation). Each folder includes the annotation (Labels.tiff), the time for each annotation (annotation_times.csv), and a video file that visualizes the annotation process. Annotations made with webKnossos, Paintera, and ITK-Snap were included. Images are free to download from: https://www.3d-cell-annotator.org/download.html
3. Neurosphere: 3D image data captured using light-sheet fluorescence microscopy, which included a spheroid with fluorescently labeled nuclei [2]. Each annotator did the manual annotation twice (manual and manual_2), while also annotating with the Minimal contour (minimal), and with the combination of Minimal contour and interpolation (manual_minimal_interpolation). Each folder includes the annotation (Labels.tiff), the time for each annotation (annotation_times.csv), and a video file that visualizes the annotation process. Images are free to download from: https://www.3d-cell-annotator.org/download.html
4. Co-culture: Multicellular spheroids were generated by using the HeLa Kyoto EGFP-alpha-tubulin/H2B-mCherry and MRC-5 fibroblast cell lines. 4 channels were acquired: DAPI-405 (ch00), Tubulin-488 (ch01), mCherry-552 (ch02), and Aktin-638 (ch03). The multi-page 16-bit Tif files have a resolution of 2048 × 2048, with pixel size of 0.14370117 µm. The distance between each image in each z-stack is 3.7 µm. To reduce the blurry effect, a deconvolution post-processing step was included only on DAPI channel provided by the LIGHTNING software (available with the LAS-X 4.4 software, Leica). Each folder includes the annotation (Labels.tiff) and the time for each annotation (annotation_times.csv). Both original images (channels) and post-processed images (LIGHTNING_DAPI_only) are included.
5. Mitotic spheroid data: Consists of 90 multicellular cancer spheroids generated from different cell lines (T-47D, 5-8F, and Huh-7D12) and imaged with a light-sheet microscope [3]. The nucleus is fluorescently labeled in the images. For each spheroid, only cells with mitotic phenotype were annotated (Labels.tiff). Images are free to download from: https://doi.org/10.6084/m9.figshare.12620078.v1
6. Melanoma_100: Benign and malignant cases were annotated by an expert in the subset of 100 images that were chosen from the HAM10000 dataset [4]. The annotations outline the lesion and separate the object from the background. Images are free to download from: https://doi.org/10.7910/DVN/DBW86T

References:

1. Saiz N. et al. Quantitative analysis of protein expression to study lineage specification in mouse preimplantation embryos. J. Vis. Exp., 108, e53654 (2016)

2. Gole L. et al. OpenSegSPIM: a user-friendly segmentation tool for SPIM data. Bioinformatics, 32, 2075–2077 (2016)

3. Diosdi A. et al. Cell lines and clearing approaches: a single-cell level 3D light-sheet fluorescence microscopy dataset of multicellular spheroids. Data in Brief Volume 36, 107090 (2021)

4. Tschandl, P. et al. The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Sci Data 5, 180161 (2018)

Napari-biomag-annotator is freely available here: https://github.com/biomag-lab/napari-biomag-annotator

All additional details about the annotation methods and image acquisition have been reported in the article cited above.

MAIN CONTACT:

Peter Horvath, Synthetic and Systems Biology Unit, Biological Research Centre (BRC) HUN-REN, 6726 Szeged, Hungary Email: horvath.peter@brc.hu

COPYRIGHT:

* Copyright (c) 2024, Akos Diosdi, Peter Horvath

* Biological Research Centre (BRC) HUN-REN, Szeged, Hungary

* All rights reserved.

*

* Redistribution and use of the material, with or without modification, is provided for academic research purposes only.

*

* This material is free; you can redistribute it and/or modify it under the terms of the GNU General Public License version 3 (or higher) as published by the Free Software Foundation.

* This material is distributed WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

* See the GNU General Public License for more details