Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
979
datasets available to search
ShareScore release 0.9.0
Dataset results
979 results for “image dataset”
CHiMP Detector Datasets: Images of Sitting Drop Protein Crystallisation Experiments with Associated Image Masks of Drops and Crystals
<p>The CHiMP Detector Datasets consist of images of protein crystallisation experiments along with corresponding zipped NumPy archive files (.npz). All images have had their histograms adjusted using the Contrast Limited Adaptive Histogram Equalization ((CLAHE) algorithm using the OpenCV library with grid size of 12 and are in JPEG format. The .npz files contain class labels and instance segmentation masks for both the experimental droplets and any crystals that an expert annotator has deemed to be interesting/mountable. To class labels and masks can be loaded in the following way:</p> <pre><code>import numpy as np # load in the mask and class label list from .npz file located at mask_path mask_file = np.load(mask_path) masks = list(mask_file["masks"].astype(int)) class_labels = list(mask_file["class_labels"])</code></pre> <p>There are two datasets within this archive:</p> <ol> <li><strong>The VMXi CHiMP Detector Dataset</strong>. This consists of 237 images of resolution 1688 × 1352 pixels with corresponding masks. These images were collected on a Rock Imager 1000 (Formulatrix, USA) automated microplate imager at the VMXi experimental facility at Diamond Light Source, UK. These images and masks were used to train the VMXi CHiMP (Crystal Hits in My Plate) Detector network that performs object detection and instance segmentation of crystals in experimental micrographs using a Mask-R-CNN architecture. The files "vmxi_detector_training.csv" and "vmxi_detector_validation.csv" provide the filenames of the members of the training and validation sets respectively.</li> <li><strong>The XChem CHiMP Detector Dataset.</strong> This consists of 350 images of resolution 1024 × 1224 pixels with corresponding masks. These images were collected on a Rock Imager 1000 (Formulatrix, USA) automated microplate imager at the Crystallisation Facility@Harwell, located in the Research Complex at Harwell (RCaH). In addition to the images in the VMXi CHiMP Detector, these images were used to train the XChem CHiMP (Crystal Hits in My Plate) Detector network that performs object detection and instance segmentation of masks and crystals in experimental micrographs using a Mask-R-CNN architecture. The files "xchem_detector_training.csv" and "xchem_detector_validation.csv" provide the filenames of the members of the training and validation sets respectively.</li> </ol>
HyperspectralBlueberries: a dataset of hyperspectral reflectance images of normal and defective blueberries
<p>The <strong>HyperspectralBluberries</strong> dataset consists of hyperspectral datacubes, which were acquired by an in-house assembled benchtop line scanning system, from 420 blueberries of two categories, including 210 sound fruit and 210 samples with various defects. The fruit samples were hand-picked from a commercial orchard. Each scanning event, which was done for an array of 42 samples, yields two files in image formats .bil (band-interleaved-by-line) and .hdr (header), which store the hyperspectral raw data and associated metadata, respectively, and are both necessary for loading hyperspectral data for processing. In addition to sample scanning, a white reference was also scanned, which can be used for standardizing spectral responses. As a result, there are 22 files in the dataset, totaling about 25 GB in file size. The sample file names are descriptive, indicating the blueberry category and number information. The dataset was used for developing machine learning models for differentiating between normal and defective blueberries, achieving an overall accuracy of 96.6%. Software programs for the modeling work are publicly available at: <a href="https://github.com/vicdxxx/Blueberry-Defect-Detection-by-Hyperspectral-Imaging">https://github.com/vicdxxx/Blueberry-Defect-Detection-by-Hyperspectral-Imaging.</a></p> <p>Details about the dataset curation and modeling experiments are described in the journal article: <a href="https://www.sciencedirect.com/science/article/pii/S2772375524000789">Deng, B., Lu, Y., Stafne, E. (2024). </a><a href="https://www.sciencedirect.com/science/article/pii/S2772375524000789">Fusing Spectral and Spatial Features of Hyperspectral Reflectance Imagery for Differentiating between Normal and Defective Blueberries. Smart Agricultural Technology</a>. <a title="Persistent link using digital object identifier" href="https://doi.org/10.1016/j.atech.2024.100473" target="_blank" rel="noreferrer noopener">https://doi.org/10.1016/j.atech.2024.100473</a>. If you use the dataset in published research, please consider citing the dataset or the <a href="https://doi.org/10.1016/j.ecoinf.2024.102546">journal article</a>. Hopefully, you find the dataset useful. </p>
GOIMAI DataSet-Phase-I of high-resolution macroscopic images for the identification of wood species
<p>GOIMAI Phase I is a wood species dataset obtained through magnifying lenses. The inputs needed in our laboratory were a mid-range mobile phone (e.g. Samsung Galaxy S7), a magnifying lens with a universal clamp attached to the lens of the mobile phone (in our case with 24x magnifier) and the wood sample itself.</p> <p>GO IMAI project - Identificación de madera por dispositivo móvil. <br>Convocatoria EI-Agri Programa Nacional de Desarrollo Rural 2014-2020 por el Ministerio de Agricultura, Pesca y Alimentación y el Fondo Europeo Agrícola de Desarrollo Rural (FEADER).</p>
Dataset of images SfM - FRM
<p>This collection of images was employed to examine the impact of various configurations in the 3D modeling process for short-distance environments. Through this analysis, settings were established to achieve submillimeter accuracy in the RMSE values of the analyzed points. The configurations evaluated included camera calibration, overlap percentage, different scale bar arrangements, and the use of both vertical and oblique images.</p>
Demo Live Cell Imaging Dataset for PetaKit5D
<p>Demo dataset for PetaKit5D (<a href="https://github.com/abcucberkeley/LLSM5DTools">https://github.com/abcucberkeley/PetaKit5D</a>). </p> <p>The dataset is for the large field of view live cell imaging of LLC-PK1 cells of nuclei and endoplasmic reticulum (ER). It contains two time points with two channels. Each time point and channel contains 4 volumetric tiles. Some demos in PetaKit5D use this dataset to demonstrate the usage.</p> <p>Please cite our paper (<a href="https://doi.org/10.1101/2023.12.31.573734">https://doi.org/10.1101/2023.12.31.573734</a>) if you use this dataset in your research:</p> <p><code>Xiongtao Ruan, Matthew Mueller, Gaoxiang Liu, Frederik Görlitz, Tian-Ming Fu, Daniel E. Milkie, Joshua L. Lillvis, Alexander Kuhn, Chu Yi Aaron Herr, Wilmene Hercule, Marc Nienhaus, Alison N. Killilea, Eric Betzig, Srigokul Upadhyayula. Image processing tools for petabyte-scale light sheet microscopy data. bioRxiv 2023.12.31.573734; doi: <a href="https://doi.org/10.1101/2023.12.31.573734">https://doi.org/10.1101/2023.12.31.573734</a></code></p>
Scanless two-photon voltage imaging example code and dataset
<p>This data and code are associated with the manuscript "Scanless two-photon voltage imaging" published in Nature Communications in June 2024 (DOI: <a href="10.1038/s41467-024-49192-2">10.1038/s41467-024-49192-2</a><span>). </span></p>
Dataset for the Julia code titled "Image Similarity Detection using ORB & Julia : Leveraging Hamming Distance & Keypoint Optimization"
<p>The above Dataset contains</p> <p>1) Real PCB Pics</p> <p>2) Real Car Fusebox pics</p> <p>3) Real & Filtered Human face pics</p> <p>4) Readme file</p> <p>The above Dataset is designed to be used for the Julia Code titled "Image Similarity Detection using ORB & Julia : Leveraging Hamming Distance & Keypoint Optimization". The Code should be run on either similar PCB pics or Car Fusebox pics or Human Face Pics(Vikrant or Mat or Paris). Proper Choice of Hamming Distance & the number of Keypoints is necessary to achieve the desired number of similar images. </p>
The IBEM Dataset: a large printed scientific image dataset for indexing and searching mathematical expressions
<p>The IBEM dataset consists of 600 documents with a total number of 8272 pages, containing 29603 isolated and 137089 embedded Mathematical Expressions (MEs). The objective of the IBEM dataset is to facilitate the indexing and searching of MEs in massive collections of STEM documents. The dataset was built by parsing the LaTeX source files of documents from the <a href="https://www.cs.cornell.edu/projects/kddcup/datasets.html">KDD Cup Collection</a>. Several experiments can be carried out with the IBEM dataset ground-truth (GT): ME detection and extraction, ME recognition, etc.</p> <p> </p> <p>The dataset consists of the following files:</p> <ul> <li>“IBEM.json”: file containing the IBEM GT information. The data is firstly organized by pages, then by the type of expression (“embedded” or “displayed”), and lastly by the GT of each individual ME. For each ME we provide: <ul> <li>xy page-level coordinates, reported as relative (%) to the width/height of the page image.</li> <li>“split” attribute indicating the number of fragments in which the ME has been split. MEs can be split over various lines, columns or pages. The LaTeX transcript of split MEs have been exactly replicated (entire LaTeX definition) for each fragment.</li> <li>“latex” original transcript as extracted from the LaTeX source files of the documents. This definition can contain user-defined macros. In order to be able to compile these expressions, each page includes the preamble of the source files containing the defined macros and the packages used by the authors of the documents.</li> <li>“latex_expand” transcript reconstructed from the output stream of the LuaLaTeX engine in which user-defined macros have been expanded. The transcript has the same visual representation as the original transcript, with the addition that the LaTeX definitions are tokenized, the order of sub/super script elements have been fixed, and matrices have been transformed to arrays.</li> <li>“latex_norm” transcript resulting from applying an extra normalization process to the “latex_expand” expression. This normalization process includes removing font information such as slant, style, and weight.</li> </ul> </li> <li>“partitions/*.lst”: files containing list of pages forming the partition sets.</li> <li>“pages/*.jpg”: individual pages extracted from the documents.</li> </ul> <p>The dataset is partitioned into various sets as provided for the ICDAR 2021 Competition on Mathematical Formula Detection. The ground-truth related to this competition, which is included in this dataset version, can also be found <a href="https://zenodo.org/record/4757865">here</a>. More information about the competition can be found in the following paper:</p> <p>D. Anitei, J.A. Sánchez, J.M. Fuentes, R. Paredes, and J.M. Benedí. ICDAR 2021 Competition on Mathematical Formula Detection. In ICDAR, pages 783–795, 2021.</p> <p> </p> <p>For ME recognition tasks, we recommend rendering the “latex_expand” version of the formulae in order to create standalone expressions that have the same visual representation as MEs found in the original documents (see attached python script “extract_GT.py”). Extracting MEs from the documents based on coordinates is more complex, as special care is needed to concatenate the fragments of split expressions. Baseline results for ME recognition tasks will soon be made available.</p>
Demo dataset for: SPACEc, a streamlined, interactive Python workflow for multiplexed image processing and analysis
<p>Multiplexed imaging technologies provide insights into complex tissue architectures. However, challenges arise due to software fragmentation with cumbersome data handoffs, inefficiencies in processing large images (8 to 40 gigabytes per image), and limited spatial analysis capabilities. To efficiently analyze multiplexed imaging data, we developed SPACEc, a scalable end-to-end Python solution, that handles image extraction, cell segmentation, and data preprocessing and incorporates machine-learning-enabled, multi-scaled, spatial analysis, operated through a user-friendly and interactive interface.</p> <p>The demonstration dataset was derived from a previous analysis and contains TMA cores from a human tonsil and tonsillitis sample that were acquired with the Akoya PhenocyclerFusion platform. The dataset can be used to test the workflow and establish it on a user's system or to familiarize oneself with the pipeline.</p>
CLIP-Embedded RedCaps Text-Image Dataset
<p>This dataset was created by applying the <a href="https://github.com/openai/CLIP">CLIP embedding</a> to the<a href="https://redcaps.xyz"> RedCaps </a>dataset. Queries are generated by OpenAI's GPT model simulating textual queries searching multimodal content, embedded via CLIP. The data was curated by Desai, Kaul, Aysola, and Johnson from data collected by Reddit, and further curated into vector data by Engels for <a href="https://arxiv.org/pdf/2402.00943">this work</a>. </p> <p>Usage of the dataset itself is subject to <a href="https://www.reddit.com/wiki/api-terms">Reddit terms</a>, <a href="https://www.redditinc.com/policies">Reddit User Agreeement, Content Policy, and Privacy Policy</a> (quoted from Desai et. al.'s <a href="https://arxiv.org/pdf/2111.11431">accompanying paper</a> for the image-and-text dataset). Usage of the <strong>queries </strong>are subject to <a href="https://openai.com/policies/row-terms-of-use/">OpenAI terms</a>. Among others, OpenAI terms prohibits using the query component of this dataset to develop models that compete against OpenAI.</p>
Demo dataset for code to quantify the 3D biofilm biovolume in images
<p>This dataset contains the raw and analyzed images that can be used as a test/demo dataset for running the code to compute the 3D biofilm biovolume. The code is stored on this Gitlab repository: https://github.com/knutdrescher/biofilm-3D-biovolume </p>
SubDiv17: A Dataset for Investigating Subjectivity in the Visual Diversification of Image Search Results
<p>This dataset facilitates the comparison of approaches aiming at the diversification of image search results. The dataset was explicitly designed for general-purpose, multi-topic queries and provides multiple ground truth annotations to allow for the exploration of the subjectivity aspect in the general task of diversification. The dataset provides images and their metadata retrieved from Flickr for around 200 complex queries. Additionally, to encourage experimentations (and cooperations) from different communities such as information and multimedia retrieval, a broad range of pre-computed descriptors is provided. The dataset was successfully validated during the MediaEval 2017 Retrieving Diverse Social Images task using 29 submitted runs. For more information, please see <a href="https://doi.org/10.1145/3204949.3208122">https://doi.org/10.1145/3204949.3208122</a>.</p>
HDRIHaven: Public domain high dynamic range image dataset
<p>High quality datasets are essential for image processing research.<br> This set contains 201 high dynamic range panoramic images (360 x 180 degrees) with jpeg previews.<br> <br> The images are originally published at HDRIHaven under CC0 license / waiver:<br> - HDRIHaven: https://hdrihaven.com/<br> <br> The images are named using the SOURCE_IDX_DESCRIPTION.EXTENSION format, e.g. <br> hdrihaven_001_ahornsteig_16k.hdr.xz to ease further reference (e.g. "hdrihaven_001").</p> <p><br> <strong>HDRIHaven:</strong><br> The photos were taken and processed by Greg Zaal, and he generously published the images under CC0 license.<br> The images were originally published at https://hdrihaven.com.</p> <p>The files have Radiance HDR type and 16K resolution (16384x8192, 134Mpixel)<br> <br> <strong>Bugfix:</strong></p> <p>One file was missing from the first version (hdrihaven_200_yellow_field_16k.hdr.xz).<br> This updated version should be used instead of the original.</p> <p> </p> <p><strong>Downloading data:</strong><br> Due to the large size, the .hdr images were compressed with xz, which is not especially fast, but has a high compression ratio. <br> <br> It could be quite time consuming to download each file individually.<br> If all of the files are required, then it might be a better approach to download only the "file_list.txt",<br> and use a download manager to download the rest of the files. E.g. on Unix systems: </p> <pre><code class="language-bash">wget -i file_list.txt </code></pre> <p><br> Direct link to the file list: <a href="https://zenodo.org/record/1285800/files/file_list.txt">https://zenodo.org/record/1285800/files/file_list.txt</a></p> <p>Alternatively, zenodo_get ( <a href="https://gitlab.com/dvolgyes/zenodo_get">http://gitlab.com/dvolgyes/zenodo_get</a> ) could be used to download the whole record, e.g.<br> </p> <pre><code>zenodo_get.py 10.5281/zenodo.1285800</code></pre> <p> </p> <p><strong>Related links:</strong></p> <p>You might also be interested in: <a href="https://doi.org/10.5281/zenodo.1245790">https://doi.org/10.5281/zenodo.1245790</a> which is a similar dataset.</p>
Thorax x-ray and CT interventional dataset for nonrigid 2D/3D image registration evaluation
<p>Thorax x-ray and CT interventional dataset for nonrigid 2D/3D image registration evaluation. Medical Physics, 2018 Nov;45(11):5343-5351. doi: 10.1002/mp.13174.</p>
AA5086 tensile tests with Portevin-Le Chatelier (PLC) effect, complete raw dataset: DIC images, raw output and elements of post-processing
<p><em>This dataset is associated with a paper published in Materials Science and Engineering: A.<br> DOI: </em><a href="https://doi.org/10.1016/j.msea.2019.01.009">10.1016/j.msea.2019.01.009</a></p> <p>Hardening and 2D digital image correlation data obtained on AA5086 sheets with Portevin-Le Chatelier (PLC) effect and Piobert-Lüder (PL) bands.</p> <ul> <li>rolled sheets</li> <li>28 tensile tests in rolling direction, room temperature</li> <li>constant imposed velocity with nominal strain rate from 1E-4 to 1E-1 Hz</li> <li>2 geometries: <ul> <li>(mac) large, Lw=60mm</li> <li>(mic or mes) small, Lw=4.2mm</li> </ul> </li> </ul> <p><strong>File contents</strong></p> <ul> <li>summary of experiments <ul> <li>libreoffice database format: experiments-summary.ods</li> <li>specifications, details, measurements and post treated quantities of each test</li> </ul> </li> <li>specimens blueprints (.pdf)</li> <li>video record of a test (.mp4)</li> <li>raw DIC images archives <ul> <li>format: <nominalStrainRate><geometry><index>.zip</li> <li>from 300 to 2300 images per test</li> </ul> </li> <li>time series <ul> <li>archived in timeSeries.zip</li> <li>format: <nominalStrainRate><geometry><index>_timeSeries.dat</li> <li>tabulation separated values, one line per picture (previous item)</li> <li>index, force, cross-head displacement, Stress, axial strain</li> </ul> </li> <li>bands kinetics post-treated quantities <ul> <li>archived in bandKinetics.zip, 2 TSV data files (one per geometry)</li> <li>stress at nucleation, strain magnitude jump and instantaneous working area relative elongation rate.</li> </ul> </li> </ul>
CODEBRIM: COncrete DEfect BRidge IMage Dataset
<p><strong>CODEBRIM: COncrete DEfect BRidge IMage Dataset</strong> for multi-target multi-class concrete defect classification in computer vision and machine learning.</p> <p>Dataset as presented and detailed in our CVPR 2019 publication: <a href="http://openaccess.thecvf.com/content_CVPR_2019/html/Mundt_Meta-Learning_Convolutional_Neural_Architectures_for_Multi-Target_Concrete_Defect_Classification_With_CVPR_2019_paper.html">http://openaccess.thecvf.com/content_CVPR_2019/html/Mundt_Meta-Learning_Convolutional_Neural_Architectures_for_Multi-Target_Concrete_Defect_Classification_With_CVPR_2019_paper.html</a> or <a href="https://arxiv.org/abs/1904.08486">https://arxiv.org/abs/1904.08486</a> . If you make use of the dataset <strong>please cite it as follows</strong>:</p> <p><strong>"Martin Mundt, Sagnik Majumder, Sreenivas Murali, Panagiotis Panetsos, Visvanathan Ramesh. <em>Meta-learning Convolutional Neural Architectures for Multi-target Concrete Defect Classification with the COncrete DEfect BRidge IMage Dataset</em>. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019"</strong></p> <p>We offer a supplementary GitHub repository with code to reproduce the paper and data loaders: <a href="https://github.com/ccc-frankfurt/meta-learning-CODEBRIM">https://github.com/ccc-frankfurt/meta-learning-CODEBRIM</a></p> <p>For ease of use we provide the dataset in multiple different versions.</p> <p>Files contained:<br> * CODEBRIM_original_images: contains the original full-resolution images and bounding box annotations<br> * CODEBRIM_cropped_dataset: contains the extracted crops/patches with corresponding class labels from the bounding boxes <br> * CODEBRIM_classification_dataset: contains the cropped patches with corresponding class labels split into training, validation and test sets for machine learning<br> * CODEBRIM_classification_balanced_dataset: similar to "CODEBRIM_classification_dataset" but with the exact replication of training images to balance the dataset in order to reproduce results obtained in the paper. </p>
Training dataset: MALDI imaging of N-glycans in murine kidney sections
<p>The files provided here are all adopted from the <a href="http://www.ebi.ac.uk/pride/archive/projects/PXD009808">PRIDE PXD009808 datasets</a> and the corresponding publication: Ove J. R. Gustafsson, Matthew T. Briggs, Mark R. Condina, Lyron J. Winderbaum, Matthias Pelzing, Shaun R. McColl, Arun V. Everest-Dass, Nicolle H. Packer, Peter Hoffmann. “MALDI imaging mass spectrometry of N-linked glycans on formalin-fixed paraffin-embedded murine kidney.” Analytical and Bioanalytical Chemistry (2015) 407: 2127. <a href="https://doi.org/10.1007/s00216-014-8293-7">https://doi.org/10.1007/s00216-014-8293-7</a></p> <p><br> Three 6µm sections of formalin-fixed paraffin-embedded murine kidney tissue specimens were prepared for MALDI imaging. To release N-linked glycans, PNGase F was printed onto two kidney sections. In the third section one area was printed with buffer to serve as a control and another area was covered with N-glycan calibrants (Gustafsson et al., Figure 4 a-c). 2,5-DHB matrix was sprayed onto the tissue sections and MALDI imaging was performed with 100 µm spatial resolution using a MALDI-TOF/TOF instrument.</p> <p><br> We processed the original imzML files to make them concise but meaningful as training data sets in the Galaxy training network (https://galaxyproject.github.io/training-material/).<br> We reduced the m/z range to 1250 – 2310 and resampled the m/z values with a step size of 0.1. The main part of the training is based on the control and first treated kidney file for which we selected representative pixels to further decrease file size (files: ‘control’, ‘treated1’). To test the results on the complete dataset we also provide a file in which both treated kidney sections, the control and the calibrant files are combined after decreasing and resampling the m/z range as described above. The combined file was normalized to the total ion current (TIC) (file: ‘all_files’). All processing steps were performed on<a href="http://https://usegalaxy.eu"> https://usegalaxy.eu</a> with the tools ‘MSI filtering’, ‘MSI combine’ and ‘MSI preprocessing’ in version 1.12.1.3).<br> Additionally, the LC-MS/MS results were extracted from table S2 of the publication by Gustafsson et al. and are provided as tabular file to enable the N-glycan identification (file: 'Glycan_IDs').</p>
(06)-He2019A-DS0001 – Tribolium castaneum foxQ2-5' line long-term live imaging dataset of embryonic development acquired with light sheet fluorescence microscopy
<p>(06)-He2019A-DS0001 – <em>Tribolium castaneum</em> foxQ2-5' line long-term live imaging dataset of embryonic development acquired with light sheet fluorescence microscopy</p>
Dataset CMKLR1-targeting peptide tracers for PET/MR imaging of breast cancer
<p>Dataset for the menuscript CMKLR1-targeting peptide tracers for PET/MR imaging of breast cancer</p>
REAVER Vascular Networks Fluorescent Image Dataset
<p><strong>Fluorescent Images of Vessel Networks from Various Murine Tissues</strong></p> <p> </p> <p><strong>Purpose</strong>: Image dataset of vascular networks with a diverse range of vessel architectures. Dataset is used to evaluate performance of several image processing programs (AngioQuant<sup>1</sup>, AngioTool<sup>2</sup>, RAVE<sup>3</sup>, REAVER). Manual analysis from ImageJ is used as ground truth to compare other programs against.</p> <ul> <li><strong>Labeling</strong>: IB4-Lectin with Alexa Flour 647</li> <li><strong>Modality</strong>: Confocal Microscope Nikon 80i CLSM</li> <li><strong>Objective</strong>: Mixture of 20x and 60x objective images</li> <li><strong>Image Format</strong>: Images originally acquired in Nikon IDS format, converted to 8-bit greyscale TIFs found in “_Original_Images” folder.</li> <li><strong>Questions</strong>: Email <a href="mailto:bac7wj@virginia.edu">bac7wj@virginia.edu</a> for inquiries.</li> </ul> <p> </p> <p><strong>External Links</strong></p> <ol> <li><strong>Manuscript</strong>:</li> <li><strong>Code repository: </strong><a href="https://github.com/bacorliss/REAVER_public">https://github.com/bacorliss/REAVER_public</a> for code to analyze this data (MATLAB 2019a).</li> </ol> <p> </p> <p><strong>Dataset Summary:</strong></p> <p>Each image folder contains 36 images. For each image:</p> <ol> <li>The first channel (red) is the segmented image with values of 0 or 255 (false or true).</li> <li>The second channel (green) is the skeleton image with values of 0 or 255 (false or true).</li> <li>The third channel (blue) is empty except for the Manual images where the third channel contains the original raw image.</li> </ol> <p> </p> <p><strong>Subfolders</strong></p> <ol> <li><strong>_Original_Images</strong>: contains raw input images.</li> <li><strong>AngioQuant_Auto</strong>: contains output images from automated analysis in AngioQuant.</li> <li><strong>AngioTool_Auto</strong>: contains output images from automated analysis in AngioTool.</li> <li><strong>ImageJ_Auto</strong>: contains output images from automated analysis in ImageJ.</li> <li><strong>ImageJ_Manual</strong>: contains output images from manual analysis in ImageJ.</li> <li><strong>RAVE_Auto</strong>: contains output images from automated analysis in RAVE.</li> <li><strong>REAVER_Auto</strong>: contains output images from automated analysis in REAVER.</li> </ol> <p> </p> <p><strong>Image Metadata and Output data</strong></p> <p>Each image folder has a .mat file called “Results.mat” containing the results of analysis in the form of the following variables all of which are 1x36 arrays (one entry for each image) unless specified otherwise:</p> <ol> <li><strong>branchpoint_RC</strong>: A 1x36 struct containing the row-column values for each branchpoint in the i<sup>th</sup> image (when organized in alphabetic order which is the order given everywhere else); Effectively the same as “BranchpointsByName.mat”</li> <li><strong>mean_diameter</strong>: The mean diameter of vessels in the image</li> <li><strong>num_branchpts</strong>: The number of branchpoints in the image</li> <li><strong>threshold_false_neg</strong>: The number of false negative pixels – a pixel is a false negative if the program has it as “false” and the manual image has the pixel as “true”</li> <li><strong>threshold_false_pos</strong>: The number of false positive pixels – a pixel is a false positive if the program has it as “true” and the manual image has the pixel as “false”</li> <li><strong>threshold_true_neg</strong>: The number of true negative pixels – a pixel is a true negative if the program has it as “false” and the manual image has the pixel as “false”</li> <li><strong>threshold_true_pos</strong>: The number of true positive pixels – a pixel is a false positive if the program has it as “true” and the manual image has the pixel as “true”</li> <li><strong>umppix</strong>: The length of the edge of one pixel in micrometers</li> <li><strong>vessel_area</strong>: The number of “true” pixels in the segmented image</li> <li><strong>vessel_length</strong>: The number of “true” pixels in the skeleton image</li> </ol> <p> </p> <p><strong>Dataset Output Data</strong></p> <p>The file “image_quantification.csv” in the base folder contains the aggregated results from each image folder. Each row contains the results for a given (Program, Image) pair. The columns are described below:</p> <ol> <li><strong>Program</strong>: Designates the program used to calculate the data for that row</li> <li><strong>Tissue_Type</strong>: Gives the tissue type for the image</li> <li><strong>Image_Name</strong>: Gives the specific name of the given image</li> <li><strong>Vessel_Length</strong>: The number of “true” pixels in the skeleton image</li> <li><strong>Vessel_Area</strong>: The number of “true” pixels in the segmented image</li> <li><strong>Mean_Diameter</strong>: The mean diameter of vessels in the image</li> <li><strong>Num_Branchpoints</strong>: The number of branchpoints in the image</li> <li><strong>Sensitivity</strong>: (Number of True Positive pixels) / (Number of True Positive pixels + Number of False Negative pixels)</li> <li><strong>Specificity</strong>: (Number of True Negative pixels) / (Number of True Negative pixels + Number of False Positive pixels)</li> <li><strong>Accuracy</strong>: (Number of True Positive pixels + Number of True Negative pixels) / (Total number of pixels)</li> <li><strong>umppix</strong>: The length of the edge of one pixel in micrometers</li> <li><strong>pix_dim</strong>: The edge length in pixels of the square image</li> </ol> <p> </p> <p><strong>References</strong></p> <p>1. Niemisto, A., Dunmire, V., Yli-Harja, O., Wei Zhang & Shmulevich, I. Robust quantification of in vitro angiogenesis through image analysis. <em>IEEE Trans. Med. Imaging</em> <strong>24</strong>, 549–553 (2005).</p> <p>2. Zudaire, E., Gambardella, L., Kurcz, C. & Vermeren, S. A Computational Tool for Quantitative Analysis of Vascular Networks. <em>PLOS ONE</em> <strong>6</strong>, e27385 (2011).</p> <p>3. Seaman, M. E., Peirce, S. M. & Kelly, K. Rapid Analysis of Vessel Elements (RAVE): A Tool for Studying Physiologic, Pathologic and Tumor Angiogenesis. <em>PLoS ONE</em> <strong>6</strong>, e20807 (2011).</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.