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.
71
datasets available to search
ShareScore release 0.9.0
Dataset results
71 results for “simulated images”
Dataset of "Neutron imaging and molecular simulation of systems from methane and p‑xylene"
<p>The dataset contains parameterizations, and input files for molecular dynamics simulations used in the study of methane dissolution in p-xylene. For selected conditions, full simulation data, i.e., trajectories and energetics are provided. All used simulation results data are provided in the table, along with the measured experimental data.</p>
Simulation of an imaging calorimeter to demonstrate GarNet on FPGA
<p>This data set is an output of a simulation of electrons and pions shot at a chunk of an imaging calorimeter. It is used in the case study for GarNet-on-FPGA, documented in <a href="https://arxiv.org/abs/2008.03601">arXiv:2008.03601</a>.</p> <p>Each HDF5 file contains the following arrays:</p> <p> Name | Shape | Description</p> <ul> <li>cluster | (10000, 128, 4) | Samples for training and inference. Outermost dimension is the event (cluster). Each cluster has maximum 128 hits, each of which has four features: x, y, z, and energy. The coordinates of the hits are in cm. The energy is in GeV. The x and y coordinates are relative to the seed hit, while the z coordinate is with respect to the calorimeter front face.</li> <li>size | (10000) | Number of hits in each cluster. The cluster array is zero-padded when the cluster size is below 128.</li> <li>truth_pid | (10000) | Identity of the primary particle (0: electron, 1: pion).</li> <li>truth_energy | (10000) | True energy of the primary particle.</li> <li>raw | (10000, 4375, 2) | Raw data (actual output of the simulation). For each event (outermost dimension), hit energy and primary fraction (innermost dimension indices 0 and 1) are given for each of the 4375 sensors. Energy is in MeV.</li> <li>coordinates | (4375, 3) | The x, y, and z coordinates of the 4375 sensors, to be used to interpret the raw data.</li> </ul> <p>See the paper for the details of the simulation.</p>
Videos of the processed microscope images and time series of the petrophysical parameters from image processing and geochemical simulation and of the measured induced polarisation [Video][Dataset]
<p>Supporting Information for the manuscript <em>Microfluidics and spectral induced polarization for direct observation and petrophysical modeling of calcite dissolution</em> published in Geophysical Research Letters</p> <ul> <li><strong>Data Set S1.</strong> Porosity, water saturation, and calcite sample perimeter from image<br>processing.</li> <li><strong>Data Set S2.</strong> Porosity, water conductivity, and pH from geochemical simulation.</li> <li><strong>Data Set S3.</strong> Real and imaginary components of the complex electrical conductivity at<br>2.5 Hz and CEC from petrophysical modeling.</li> <li><strong>Movie S1.</strong> Dissolution of the calcite sample with the detected contour superimposed in<br>white on the grayscale images. Time, length scale, and flow direction are indicated. In<br>case of problems launching the file, we recommend using VLC Media Player software.</li> <li><strong>Movie S2.</strong> Segmented images of the CO2 bubbles produced by the calcite dissolution.<br>Time, length scale, and flow direction are indicated. In case of problems launching the<br>file, we recommend using VLC Media Player software.</li> </ul>
Pre-training with simulated ultrasound images for breast mass segmentation and classification - dataset
<p>Dataset assosiated with the MICCAI Workshop on Data Engineering in Medical Imaging paper: "Pre-training with Simulated Ultrasound Images for Breast Mass Segmentation and Classification"</p>
Scanned images of monocultures and mixtures of six grassland plant species roots, and of simulated fine roots
<p>Soil core samples were taken from a multi-species grassland experiment with field plots of monocultures and mixtures of six grassland plant species: <em>Lolium perenne</em> L. (PRG),<em> Phleum pratense</em> L. (TIM), <em>Trifolium pratense</em> L. (RC), <em>Trifolium repens</em> L. (WC), <em>Cichorium intybus </em>L. (CHIC), and <em>Plantago lanceolata </em>L.. The multi-species plots had a two species mixture with <em>Trifolium repens </em>L. and<em> Lolium perenne</em> L. (PRGWC), and a 6 species mixture with all species mentioned above. The cores were separated into soil depths of 0-10 cm, 10-15 cm and 15-20 cm and the roots separated from the soil.</p> <p>A ground-truth image set was created to simulate fine roots using fishing line. The fishing line used was a clear copolymer monofilament (Greys<sup>TM</sup> Greylon Tippet Material 3 lb), measured using a scanning electron microscope (Hitachi SU8200) to be 0.14 mm in diameter. The fishing line was used in its clear colour or coloured black using a permanent marker to simulate unstained and stained fine roots respectively. The fishing line was cut into lengths of 30 cm or 5 cm. </p> <p>Roots and fishing line were scanned using an Epson Perfection V800 flatbed scanner at 600 dpi. </p> <p>The Roots ZIP file contains a folder for the scanned root images and the Line zip file contains a folder with the scanned fishing line. The excel spreadsheet describes the naming convention for the images.</p> <p>Further details about the root sampling and image acquisition can be found in the publication that analyses these images: <a href="https://doi.org/10.1002/ppj2.20034">https://doi.org/10.1002/ppj2.20034</a></p>
Aegean v2.0 Simulated Test Image
<p>A radio sky image simulated for the purposes of testing v2.0 of the Aegean source finding algorithm.</p>
Simulated dMRI images and ground truth of random fiber phantoms in various configurations
<p>This archive contains simulated dMRI images of random fiber phantoms in various configurations created with Fiberfox and other tools available in MITK Diffusion (<a href="http://mitk.org/wiki/DiffusionImaging">http://mitk.org/wiki/DiffusionImaging</a>). RandomFibers_Example.png illustrates one of the random fiber configurations used for these phantoms.</p> <p>If you are using any of these datasets or the tools used to generate them, please don't forget to cite the dataset itself as well as other relevant publications.</p> <p>Each subfolder contains the following elements:<br> The simulated dMRI image with b-values and gradient directions: dwi.nii.gz, dwi.bvals, dwi.bvecs<br> The fibers used for simulation: AllBundles.fib (binary vtk format)<br> parameters.ffp: Fiberfox simulation parameters<br> parameters.ffp.bvals: b-value file for Fiberfox simulation<br> parameters.ffp.bvecs: gradient vector file for Fiberfox simulation<br> parameters.ffp_VOLUME1.nii.gz: fiber compartment volume fraction map for Fiberfox simulation<br> The logfile detailing all steps of the generation process of the respective phantom: LOGFILE.json</p> <p>bundles: folder containing the individual fiber bundles (binary vtk format .fib)<br> centroids: folder containing the centerlines of each bundle<br> masks: folder containing the binary envelope of each bundle<br> peaks: folder containing the principal fiber direction image (peaks) of each bundle</p> <p>Each subfolder contains the fibers and dMRI simulations with the following fiber specifications:<br> Phantom 1:<br> - Number of bundles: 25<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 2:<br> - Number of bundles: 25<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 3:<br> - Number of bundles: 25<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 4:<br> - Number of bundles: 25<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 5:<br> - Number of bundles: 25<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 6:<br> - Number of bundles: 25<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 7:<br> - Number of bundles: 25<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 8:<br> - Number of bundles: 25<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 9:<br> - Number of bundles: 50<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 10:<br> - Number of bundles: 50<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 11:<br> - Number of bundles: 50<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 12:<br> - Number of bundles: 50<br> - Fiber density: 250 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 13:<br> - Number of bundles: 50<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 14:<br> - Number of bundles: 50<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 0-30 in degree<br> - Bundle start radius: 15-30 in mm</p> <p>Phantom 15:<br> - Number of bundles: 50<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 5-15 in mm</p> <p>Phantom 16:<br> - Number of bundles: 50<br> - Fiber density: 50-500 streamlines per cm²<br> - Bundle curvature: 30-60 in degree<br> - Bundle start radius: 15-30 in mm</p>
Dataset for the publication "Superconducting gravimeter observations show that satellite-derived snow depth image improves the simulation of the snow water equivalent evolution in a high alpine site"
<p>This datasset contains data to reproduce the following figures of the paper <em>Superconducting gravimeter observations show that satellite-derived snow depth image improves the simulation of the snow water equivalent evolution in a high alpine site</em>:</p> <ul> <li> <p>Time series data of Figures 1c and 2</p> </li> <li> <p>Data (*.asc) used for plotting Figures 1d and 1e (as well as Figure S3 and S4)</p> </li> <li>Pléiades snow depth map (Figure S1)</li> <li> <p>Data used for plotting Figure S2</p> </li> </ul> <p> </p>
Polymer simulations guide the detection and quantification of chromatin loop extrusion by imaging
<p><strong>Dataset description</strong></p> <p>This dataset contains coordinates of the two anchors of a 150 kb simulated loop (static imaging) and anchor-anchor distances of a 150 kb loop in simulations where the extruder was allowed to unbind from the polymer (dynamic imaging).</p> <p><strong>Dataset description - Static imaging</strong></p> <p>This dataset contains coordinates of the two anchors of a 150 kb simulated loop.</p> <p>The subfolder 'Free' contains coordinates from a polymer not submitted to loop extrusion (1,000 independent simulations). The subfolder 'Loop' contains loop anchor coordinates in a polymer submitted to loop extrusion (4,000 independent simulations). In simulations with extrusion, the polymer chain was simulated such as it went through 3 different states : i) Open state (absence of loops) ii) Extruding state where the loop size increases with time (anchor-anchor distance decreases) and iii) Closed state corresponding to a stable loop with the two anchors in contact.</p> <p> </p> <p><strong>Structure of data - Static imaging</strong></p> <p>In each .txt file, the first three columns correspond to the XYZ coordinates of the anchor (in µm), the fourth column indicates the state label (0=Open, 1=Extruding, 2=Closed). Each row corresponds to a simulation timepoint (2991 timepoints in polymers submitted to loop extrusion).</p> <p>The simulation ID is indicated at the end of each .txt file.</p> <p>The two anchors of the loop are bead #275 and bead #324.</p> <p> </p> <p><strong>Structure of data - Dynamic imaging</strong></p> <p>Each .txt file contains the anchor-anchor distance (in µm) as function of time for approximately 10,000 independent simulations. Each column is an independent simulation. Each row is a simulation timepoint (0.3 s / simulation unit). The different .txt files correspond to different localization errors, indicated in µm in XY and Z.</p> <p>'list_deb_closed_10000.p' is a dictionary whose keys are the simulation ID (corresponding to the columns of the .txt files), and the values are the frame at which extrusion begins.</p> <p>'list_end_closed_10000.p' is a dictionary whose keys are the simulation ID (corresponding to the columns of the .txt files), and the values are the frame at which extrusion ends.</p>
Simulated water models with Apoferritin for use in cryo-EM image simulations with amorphous ice
<p>This dataset contains the atomic coordinates of several water models produced using the NAMD molecular dynamics software. The contents of each file is listed below.</p> <ul> <li><strong><em>water_81_coords.pdb</em></strong> - water only in a cubic box with side length 81A</li> <li><strong><em>water_243_coords.pdb</em></strong> - water only in a cubic box with side length 243A</li> <li><strong><em>water_486_coords.pdb</em></strong> - water only in a cubic box with side length 486A</li> <li><strong>water_567_coords.pdb</strong> - water only in a cubic box with side length 567A</li> <li><strong><em>water_645_coords.pdb</em></strong> - water only in a cubic box with side length 645A</li> <li><strong><em>water_735_coords.pdb</em></strong> - water only in a cubic box with side length 735A</li> <li><strong><em>apo_water_723_coords.pdb</em></strong> - water and apoferritin in a cubic box with side length 723A where apoferritin atoms are constrained</li> </ul>
Simulation and data analysis for "Center-of-Mass Corrections in Associated Particle Imaging"
<p>IPython notebook used to run simulations and generate all plots the paper "Center-of-Mass Corrections in Associated Particle Imaging". The notebook also includes some more analysis and plots not shown in the paper.</p>
Reproduction packages for the paper "Spectral and Imaging properties of Sgr A∗ from High-Resolution 3 DGRMHD Simulations with Radiative Cooling"
<p>This is a basic reproduction package for the paper"Spectral and Imaging properties of Sgr A∗ from High-Resolution 3D GRMHD Simulations with Radiative Cooling" by Yoon et al. (2020). It aims to provide the most important data products to check and reproduce the main results of the paper.</p>
Real and simulated cross-sectional and longitudinal images of hair
<p>This is the dataset containing simulated and real data used in the analyses for the paper "<em><a href="https://www.biorxiv.org/content/10.1101/2020.11.24.392191v1">High-throughput phenotyping methods for quantifying hair fiber morphology</a>" </em>and is part of the<em> <a href="https://tinalasisi.github.io/2020_HairPheno_manuscript/index.html">Hair Phenotyping Methods Project</a> </em>run by <a href="https://www.tinalasisi.com/">Tina Lasisi</a>. </p> <p>The data can be analyzed with the <em>fibermorph</em> Python package available on <a href="https://pypi.org/project/fibermorph/">PyPi</a> and <a href="https://github.com/tinalasisi/fibermorph">Github</a>. </p> <p>This repository has 2 datasets with 2 different types of data:</p> <ol> <li>Simulated hair data <ol> <li>Cross-sectional data (simulated ellipses)</li> <li>Curvature data (simulated arcs)</li> </ol> </li> <li>Real hair data <ol> <li>Cross-sectional data (micrographs of hair fiber cross-sections)</li> <li>Curvature data (longitudinal images of hair fiber fragments)</li> </ol> </li> </ol> <p>Visit the <a href="https://tinalasisi.github.io/2020_HairPheno_manuscript/index.html"><em>Hair Phenotyping Methods Project</em></a> website for the most up to date information about this project and any updates relevant to this dataset.</p> <p>-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------</p> <p><strong>Details</strong></p> <p><strong>Simulated data</strong></p> <ol> <li>Cross-sectional data <ul> <li>These ellipses were simulated with a python script developed as part of <a href="https://github.com/tinalasisi/fibermorph/"><em>fibemorph</em></a>. A version of that code that doesn't require the original Python package has been made available with the dataset (sim_ellipse.py).</li> <li>The script simulates a single cross-section per image.</li> <li>Each image has a width of 5200px and a height of 3900 with a resolution set to 4.25 px/micron.</li> </ul> </li> <li>Curvature data <ul> <li>An R script used for curvature simulation, written by <a href="https://www.arslanzaidi.com/">Arslan Zaidi,</a> has also been made available with this dataset (sim_curvature.R). </li> <li>The script generates 25 arcs per image. We used a set length of 1.57.</li> <li>Each image has a resolution of 132 px/mm. </li> </ul> </li> </ol> <p>Please note that due to the use of random generations, it is not possible to recreate the exact same datasets that are saved here. </p> <p><strong>Real data</strong></p> <p>The real data images are very large files and have been split into multiple zip files. Please check the specific instructions for unzipping split zip files for your OS. </p> <p>The images are from hair samples collected by the <a href="https://shriverlab.psu.edu/">Shriver Lab</a> at Penn State. There were a total of 192 samples, although not all images made it past quality control so certain IDs may have cross-section images but not curvature images or vice versa. </p> <p>The images have been de-identified and the hair samples for these individuals were collected with informed consent and ethical approval by The Pennsylvania State University Institutional Review Board (#44929 and #45727).</p> <ol> <li>Cross-sectional data <ol> <li>We developed and used <a href="https://www.protocols.io/view/sample-preparation-protocol-for-cross-sectional-mi-bbwcipaw">this protocol</a> to embed, section, and image the hairs.</li> <li>We embedded 6 samples per person and took images of both sides of the sectioned sample (A and B). These should be mirror images of each other.</li> </ol> </li> <li>Curvature data <ol> <li>We developed and used <a href="https://www.protocols.io/view/sample-preparation-protocol-for-hair-fiber-curvatu-bbweipbe">this protocol</a> to cut, wash, and image the hairs.</li> <li>We used 3-5 hairs per person where available. A number of samples did not have enough hair for this, so the images contain fewer fragments. We have made these available for full transparency although we filtered them from our analyses downstream.</li> </ol> </li> </ol> <p>Please see the <a href="https://github.com/tinalasisi/2020_HairPheno_manuscript">GitHub repository</a> for additional related participant data we used in our analyses.</p> <p> </p> <p> </p>
Library of simulated root images
<p>This depository contains the images and associated RSML files used in the paper entitled "Using a structural root system model for an in-depth assessment of root image analysis pipeline" from the same authors.</p> <p> </p> <p>It contains: </p> <p>- 10.000 simulated root images;</p> <p>- 10.000 corresponding RSML files;</p> <p>- .csv files containing the ground-truth data for each modelled root system;</p> <p>- .csv files containing the image descriptors extracted using RIA-J;</p> <p>- .tps files containing the shape descriptors data.</p> <p> </p> <p>The codes used with this data is available here: https://zenodo.org/record/62063</p>
Library of simulated root images, with different noise levels
<p>This depository contains the images and associated RSML files used in the paper entitled "Using a structural root system model to evaluate and improve the accuracy of root image analysis pipelines" from the same authors.</p> <p> </p> <p>It contains: </p> <p>- 30.000 simulated root images of 10.000 different root systems, with different noise levels (0=null, 1=medium, 3=high);</p> <p>- 10.000 corresponding RSML files;</p> <p>- .csv files containing the ground-truth data for each modelled root system (500-data.csv);</p> <p>- .csv files containing the image descriptors extracted using RIA-J (500-descriptors.csv);</p> <p> </p> <p>The codes used to generate and analyse this dataset is available here: http://doi.org/10.5281/zenodo.208499</p> <p>Note about the metrics computed from the RSML files (contained in 500-data.csv): the root systems were simulated in a constrained 2D space (rhizotron-like). Therefore, the RSML_reader plugin computed the lengths of the different roots in 2D only (despite the fact that a Z coordinate is present in the the RSML files). </p> <p> </p> <p><strong>CORRECTION</strong>: There is a scale issue in the 500-descriptors.csv file. "area" and "convexhull" columns values should be divided by 10 </p> <p> </p> <p> </p>
An urban traffic dataset composed of visible images and their semantic segmentation generated by the CARLA simulator
<p><strong>If you use this dataset please cite this paper: Rosende, S.B.; Gavilán, D.S.J.; Fernández-Andrés, J.; Sánchez-Soriano, J. An Urban Traffic Dataset Composed of Visible Images and Their Semantic Segmentation Generated by the CARLA Simulator. <em>Data</em> 2024, <em>9</em>, 4. <a href="https://doi.org/10.3390/data9010004">https://doi.org/10.3390/data9010004</a></strong></p> <p>A dataset of aerial urban traffic images and their semantic segmentation is presented to be used to train computer vision algorithms, among which those based on convolutional neural networks stand out. The images have been generated using the CARLA simulator (but would be like those that could be obtained with fixed aerial cameras or by using AUVs) in the field of intelligent transportation management. The presented dataset is available and accessible to improve the performance of vision and road traffic management systems, especially for the detection of incorrect or dangerous maneuvers.</p>
Investigating the Unsteady Dynamics of a Multi-Jet Impingement Cooling Flow Using Large Eddy Simulation - Promotional Video and Image
<p>Video: Volume rendering of temperature field obtained from a large eddy simulation of 9 inline impinging jets in a narrow channel.</p> <p>Image: Turbulent structures as isosurface of Q-criterion with an illustration of laser sheet used for PIV measurement.</p> <p>The results were presented at the ASME Turbo Expo 2024 (paper number GT2024-122465) and published in the ASME Journal of Turbomachinery (<a href="https://doi.org/10.1115/1.4066508">https://doi.org/10.1115/1.4066508</a>). The accepted manuscript of the paper is available under <a href="https://elib.dlr.de/207257/">https://elib.dlr.de/207257/</a>.</p> <p>The simulations were performed on DLR's HPC system <a href="https://www.dlr.de/en/research-and-transfer/research-infrastructure/hpc-cluster/cara">CARA</a> within the DLR project InnoCool.</p>
Digital Pediatric Image Quality Phantoms and Simulations
<p>Pediatric IQ Phantoms is a dataset of virtual phantoms and their computed tomography (CT) images for assessing the image quality of image-based CT denoising products in pediatric patients. The phantoms in the dataset enable assessment of <a title="IEC Standard 61223-5-3" href="https://webstore.iec.ch/publication/59789" target="_blank" rel="noopener">standard measures of image quality</a> – CT number accuracy, noise magnitude, uniformity, contrast dependent spatial resolution, and low contrast detectability. These phantoms are designed to span the range of pediatric effective waist diameters (Table 1) and thus can be used to assess pediatric image quality performance.</p> Table 1: Pediatric subgroups investigated defined by <table><tbody> <tr> <td><strong>Subgroup</strong></td> <td><strong>Age Range</strong></td> <td><strong>Waist Diameter Range</strong></td> </tr> <tr> <td>Newborn</td> <td>≤ 1 mo</td> <td>≤ 11.5 cm</td> </tr> <tr> <td>Infant</td> <td>> 1 mo & < 2 yrs</td> <td>> 11.5 cm & ≤ 16.8 cm</td> </tr> <tr> <td>Child</td> <td>> 2 yrs & ≤ 12 yrs</td> <td>> 16.8 cm & ≤ 23.2 cm</td> </tr> <tr> <td>Adolescent</td> <td>> 12 yrs & < 21 yrs</td> <td>> 23.2 cm & < 34 cm</td> </tr> <tr> <td>Adult</td> <td>≥ 22 yrs</td> <td>≥ 34 cm</td> </tr> </tbody> </table> <p>These phantoms include:</p> <ul> <li>CTP404 multi-contrast phantom: for assessing CT number accuracy and contrast-dependent spatial resolution. <ul> <li>CTP404 is a modified version of the sensitometry module CTP404 from the Catphan 600 phantom (The Phantom Laboratory, Salem, NY).</li> <li>This cylindrical phantom has eight unique contrast inserts ranging from -1000 to +900 HU in a uniform background of 0 HU. In its standard size, CTP404 has a diameter of 150 mm with 12 mm diameter inserts. Due to the sharp intersection between the phantom background and multi-contrast inserts, this module was used to evaluate contrast-dependent image sharpness using the contrast-dependent modulation transfer function.2</li> </ul> </li> <li>MITA LCD phantom: for assessing low contrast detectability. <ul> <li>The MITA-LCD phantom is a cylindrical phantom filled with water-equivalent attenuation material and contains four low-contrast disk inserts of different size and contrast combinations: 3mm-14HU, 5mm-7HU, 7mm-5HU, 10mm-3HU.3,4</li> </ul> </li> <li>Uniform water phantom: for assessing noise and noise texture. <ul> <li>The uniform water phantom is a cylindrical phantom filled with water-equivalent attenuation material.</li> </ul> </li> </ul>
MedMNIST-C: Comprehensive benchmark and improved classifier robustness by simulating realistic image corruptions
<p><strong>Abstract: </strong>The integration of neural-network-based systems into clinical practice is limited by challenges related to domain generalization and robustness. The computer vision community established benchmarks such as ImageNet-C as a fundamental prerequisite to measure progress towards those challenges. Similar datasets are largely absent in the medical imaging community which lacks a comprehensive benchmark that spans across imaging modalities and applications. To address this gap, we create and open-source MedMNIST-C, a benchmark dataset based on the MedMNIST+ collection, covering 12 datasets and 9 imaging modalities. We simulate task and modality-specific image corruptions of varying severity to comprehensively evaluate the robustness of established algorithms against real-world artifacts and distribution shifts. We further provide quantitative evidence that our simple-to-use artificial corruptions allow for highly performant, lightweight data augmentation to enhance model robustness. Unlike traditional, generic augmentation strategies, our approach leverages domain knowledge, exhibiting significantly higher robustness when compared to widely adopted methods. By introducing MedMNIST-C and open-sourcing the corresponding library allowing for targeted data augmentations, we contribute to the development of increasingly robust methods tailored to the challenges of medical imaging. The code is available at <a href="https://github.com/francescodisalvo05/medmnistc-api">github.com/francescodisalvo05/medmnistc-api</a>.</p> <blockquote> <p>This work has been accepted at the Workshop on Advancing Data Solutions in Medical Imaging AI @ MICCAI 2024 [<a href="https://arxiv.org/pdf/2406.17536" target="_blank" rel="noopener">preprint</a>].</p> </blockquote> <p><strong>Note: </strong>Due to space constraints, we have uploaded all datasets except TissueMNIST-C. However, it can be reproduced via our APIs. </p> <p><strong>Usage: </strong>We recommend using the demo code and tutorials available on our GitHub <a href="https://github.com/francescodisalvo05/medmnistc-api">repository</a>.</p> <p><strong>Citation: </strong>If you find this work useful, please consider citing us:</p> <blockquote> <pre>@article{disalvo2024medmnist, title={MedMNIST-C: Comprehensive benchmark and improved classifier robustness by simulating realistic image corruptions}, author={Di Salvo, Francesco and Doerrich, Sebastian and Ledig, Christian}, journal={arXiv preprint arXiv:2406.17536}, year={2024} }</pre> </blockquote> <p><strong>Disclaimer: </strong>This repository is inspired by MedMNIST APIs and the ImageNet-C repository. Thus, please also consider citing <a href="https://www.nature.com/articles/s41597-022-01721-8" rel="nofollow">MedMNIST</a>, the respective source datasets (described <a href="https://medmnist.com/" rel="nofollow">here</a>), and <a href="https://arxiv.org/abs/1903.12261" rel="nofollow">ImageNet-C</a>.</p>
BRAIN Journal-ANNSVM: A Novel Method for Graph-Type Classification by Utilization of Fourier Transformation, Wavelet Transformation, and Hough Transformation-Figure 8. Simulation of Coiflet 1 (PyWavelets discussion group, 2008), analyzing as one-dimensional images
<p>Using only the wavelet coefficients was inadequate for classification. For example, for the pie chart, we obtained large wavelet coefficients located in the low-frequency domain; however, if we changed a circle in the pie chart to other shapes, such as a radar chart, the wavelet transformation gave results that were similar to those of the original pie chart. The Hough transformation can solve this problem since it detects the shapes of objects</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.