Find research datasets worth reusing

restrictedApr 2022View details →

International Virtual Permeability Benchmark 3D Image Dataset of the Fiber Tow Microscopic Sample

This microscopic 3D sample representing a part of the fiber tow of the woven composite reinforcement was used in the first phase of the international Virtual permeability benchmark. Its results are reported in [1-3]. While the permeability had been investigated before for various porous media, this benchmark exercise was the first contribution in the field of real fibrous microstructures. The 3D image was extracted from a 3D X-ray microscope scan of a composite sample prepared at Leibniz-Institut für Verbundwerkstoffe GmbH (IVW, Kaiserslautern, Germany). The segmentation of the 3D image was performed at the Research Institute in Civil Engineering and Mechanics (GeM) of Ecole Centrale de Nantes (Nantes, France).   The segmented binary volume image data (.raw file) has the following characteristics: <ul> <li>dimensions of 1003 x 124 x 973 voxels;</li> <li>0.5213 µm3 voxel size;</li> <li>8-bit;</li> <li>little-endian byte order.</li> </ul> The provided corresponding non-segmented volume image has the following characteristics: <ul> <li>dimensions of 1003 x 124 x 973 voxels;</li> <li>0.5213 µm3 voxel size;</li> <li>16-bit;</li> <li>little-endian byte order.</li> </ul> The microscopic volume contains about 400 fibers of a single tow and covers the entire tow through-thickness dimension. Its average fiber volume fraction is 56.5%, though it varies locally along the fiber direction. The structure of the tow represents 3 yarns twisted together. The tow originates from the glass twill-weave fabric HexForce 01102 by HEXCEL.   [1]    Elena Syerko, Tim Schmidt, David May, Christophe Binetruy, Suresh G. Advani et al. Benchmark Exercise on Image-Based Permeability Determination of Engineering Textiles: Microscale Predictions // Composites Part A: Applied Science and Manufacturing – 2023: 107397. [2]    Elena Syerko, Christophe Binetruy, Tim Schmidt, David May. Micro-Scale Results from the Benchmark Exercise on the Image-Based Permeability Prediction of Composite Reinforcements // Proceedings of IBSim-4i – Image-Based Simulation for Industry 2022, London, UK – 2022. [3]    David May, Elena Syerko, Tim Schmidt, Christophe Binetruy, Luisa Rocha da Silva, Stepan Lomov, Suresh Advani. Benchmarking Virtual Permeability Predictions of Real Fibrous Microstructure // Proceedings of the American Society for Composites – 36th Technical Conference, Texas, USA – 2021.

restrictedJul 2022View details →

Dataset related to article "Diffusion weighted magnetic resonance imaging for kidney cyst volume quantification and non-cystic tissue characterization in ADPKD"

Kidney volumes, demographic features, and median DWI-based parameters related to the individual patients and healthy volunteers included in the study

restrictedApr 2018View details →

Data for publication: Weighted manifold alignment using wave kernel signatures for aligning medical image datasets

2D dynamic sagittal MR images, corresponding to volunteers E-H in the publication "Weighted manifold alignment using wave kernel signatures for aligning medical image datasets".

restrictedSep 2023View details →

Multimodal OCTA and Fundus Image dataset for detection of Diabetic Retinopathy

This is a Multimodal OCTA and Fundus Image dataset curated for detection of Diabetic Retinopathy at Natasha Eye care and Research Institute. In all 224 images from 74 persons were captured in Version-1 and 338 images from 111 person in Version-2 Two modalities for every individual were taken simultaneously using Eidon camera for Fundus images and Optovue Avanti edition for OCTA images. we have three categories of classification in this dataset         1. NO DR   2. MILD DR  3. MODERATE DR   CONSENT Forms are taken for every Individual.

restrictedJul 2021View details →

Dataset to submitted manuscript "Imaging the blinking properties of a mitochondrial fluorophore marker turns it into a multi-functional sensing probe"

This folder contains all raw data underlying the results presented in a manuscript, submitted to Light: Science & Applications, and entitled:   Imaging the blinking properties of a mitochondrial fluorophore marker turns it into a multi-functional sensing probe   Authored by: Zhixue Dua, Joachim Pigueta,+, Gleb Baryshnikovb,+, Johan Tornmalma, Baris Demirbaya, Hans Ågrenb, Jerker Widengrena,*   a Royal Institute of Technology (KTH), Experimental Biomolecular Physics, Dept. Applied Physics, Albanova Univ Center 106 91 Stockholm, Sweden b Royal Institute of Technology (KTH), Dept Theroretical Chemistry and Biology, Albanova Univ Center 106 91 Stockholm, Sweden + Contributed equally * Corresponding author: Email:  <a href="mailto:jwideng@kth.se">jwideng@kth.se</a>, Phone: +46-8-7907813   The data files are grouped into the different techniques used to generate them, and refer to the figures/tables in the manuscript where the extracted results are presented.   ABSTRACT By transient state (TRAST) measurements, monitoring the average fluorescence intensity upon excitation modulation, we show that the mitochondrial localization fluorophore 10-Nonyl Acridine Orange (NAO) exhibits prominent singlet-triplet state transitions and can act as a light-induced Lewis acid forming a red-emissive doublet radical. The blinking properties resulting from these transitions can be monitored in a broadly applicable manner, and under biologically relevant conditions. By TRAST studies of NAO in small unilamellar vesicles (SUVs) we show that these blinking properties are highly environment sensitive, specifically reflecting local oxygen concentrations, redox conditions, membrane charge, fluidity and lipid compositions. In SUVs containing the phospholipid cardiolipin (CL) the NAO blinking properties depend not only on the concentration of the CL added, but also on the CL acyl chain composition. The blinking also reflects hydroxyl ion dependent transitions to and from the NAO doublet radical. This makes it possible to monitor local pH and buffering properties, from a 3D bulk buffer above the membrane, or from a 2D buffer at the membrane surface itself. Finally, by live cell TRAST imaging, we show that the fluorescence blinking properties of NAO can also be imaged in a spatially resolved manner, and that this commercially available location-specific fluorophore thereby can be turned into a multi-parametric intracellular sensing probe. This study demonstrates new possibilities for fundamental membrane studies in artificial vesicles and live cells, using existing fluorophore markers, monitoring parameters and conditions of large biological relevance, which are difficult to retrieve by other means.

restrictedAug 2021View details →

Dataset related to article "Automated Head Tissue Modelling Based on Structural Magnetic Resonance Images for Electroencephalographic Source Reconstruction"

SCORING SEGMENTATIONS <ul> <li>Qualitative segmentation scores by two raters (rater1; rater2).</li> <li>Scale: excellent (4); good (3); doubtful (2) and failed (1).</li> </ul>   DATABASES <ul> <li>IXI database, Imperial College of London (<a href="https://brain-development.org/ixi-dataset/">https://brain-development.org/ixi-dataset/</a>)</li> <li>Autism Brain Imaging Data Exchange (ABIDE) database (<a href="http://fcon_1000.projects.nitrc.org">http://fcon_1000.projects.nitrc.org</a>)</li> <li>SchizConnect database (<a href="http://schizconnect.org">http://schizconnect.org</a>)</li> </ul>   SEGMENTATION METHODS <ul> <li>MR-TIM (Taberna et al., 2021), green rows</li> <li>WTS (Liu et al., 2017), red rows</li> </ul>   TABLES IXI_young <ul> <li>20 MRI from the IXI database, participants 20–35 years old;</li> <li>MR scanners: Philips Intera 3.0T (HH); Philips Gyroscan Intera 1.5T (G)</li> </ul> IXI_older <ul> <li>20 MRI from the IXI database, participants 60–75 years old;</li> <li>MR scanners: Philips Intera 3.0T (HH); Philips Gyroscan Intera 1.5T (G)</li> </ul> ABIDE <ul> <li>10 MRI from the ABIDE database, participants 18-25 years old;</li> <li>MR scanner: Philips Achieva 3.0T</li> </ul> SchizConnect <ul> <li>10 MRI from the SchizConnect database, participants 19-66 years old;</li> <li>MR scanner: Siemens Trio Tim 3.0T</li> </ul>   REFERENCES Liu, Q., Farahibozorg, S., Porcaro, C., Wenderoth, N., & Mantini, D. (2017). Detecting large-scale networks in the human brain using high-density electroencephalography. Hum Brain Mapp, 38(9), 4631-4643. doi:10.1002/hbm.23688 Taberna, G. A., Samogin, J., & Mantini, D. (2021). Automated Head Tissue Modelling Based on Structural Magnetic Resonance Images for Electroencephalographic Source Reconstruction. Neuroinformatics. doi:10.1007/s12021-020-09504-5

restrictedDec 2022View details →

Dataset related to the article "Cardiovascular magnetic resonance images with susceptibility artifacts: artificial intelligence with spatial-attention for ventricular volumes and mass assessment"

This record contains raw data related to the article "Cardiovascular magnetic resonance images with susceptibility artifacts: artificial intelligence with spatial-attention for ventricular volumes and mass assessment" Abstract Background Segmentation of cardiovascular magnetic resonance (CMR) images is an essential step for evaluating dimensional and functional ventricular parameters as ejection fraction (EF) but may be limited by artifacts, which represent the major challenge to automatically derive clinical information. The aim of this study is to investigate the accuracy of a deep learning (DL) approach for automatic segmentation of cardiac structures from CMR images characterized by magnetic susceptibility artifact in patient with cardiac implanted electronic devices (CIED). Methods In this retrospective study, 230 patients (100 with CIED) who underwent clinically indicated CMR were used to developed and test a DL model. A novel convolutional neural network was proposed to extract the left ventricle (LV) and right (RV) ventricle endocardium and LV epicardium. In order to perform a successful segmentation, it is important the network learns to identify salient image regions even during local magnetic field inhomogeneities. The proposed network takes advantage from a spatial attention module to selectively process the most relevant information and focus on the structures of interest. To improve segmentation, especially for images with artifacts, multiple loss functions were minimized in unison. Segmentation results were assessed against manual tracings and commercial CMR analysis software cvi42(Circle Cardiovascular Imaging, Calgary, Alberta, Canada). An external dataset of 56 patients with CIED was used to assess model generalizability. Results In the internal datasets, on image with artifacts, the median Dice coefficients for end-diastolic LV cavity, LV myocardium and RV cavity, were 0.93, 0.77 and 0.87 and 0.91, 0.82, and 0.83 in end-systole, respectively. The proposed method reached higher segmentation accuracy than commercial software, with performance comparable to expert inter-observer variability (bias ± 95%LoA): LVEF 1 ± 8% vs 3 ± 9%, RVEF − 2 ± 15% vs 3 ± 21%. In the external cohort, EF well correlated with manual tracing (intraclass correlation coefficient: LVEF 0.98, RVEF 0.93). The automatic approach was significant faster than manual segmentation in providing cardiac parameters (approximately 1.5 s vs 450 s). Conclusions Experimental results show that the proposed method reached promising performance in cardiac segmentation from CMR images with susceptibility artifacts and alleviates time consuming expert physician contour segmentation.

restrictedJan 2023View details →

Dataset related to article "Frozen Section Analysis and Real-Time Magnetic Resonance Imaging of Surgical Specimen Oriented on 3D Printed Tongue Model to Assess Surgical Margins in Oral Tongue Carcinoma: Preliminary Results"

This record contains raw data related to article “Frozen Section Analysis and Real-Time Magnetic Resonance Imaging of Surgical Specimen Oriented on 3D Printed Tongue Model to Assess Surgical Margins in Oral Tongue Carcinoma: Preliminary Results" Abstract Background: A surgical margin is the apparently healthy tissue around a tumor which has been removed. In oral cavity carcinoma, a negative margin is considered ≥ 5 mm, a close margin between 1 and 5 mm, and a positive margin ≤ 1 mm. Currently, the intraoperative surgical margin status is based on the visual inspection and tissue palpation by the surgeon and intraoperative histopathological assessment of the resection margins by frozen section analysis (FSA). FSA technique is limited and susceptible to sampling errors. Definitive information on the deep resection margins requires postoperative histopathological analysis. Methods: We described a novel approach for the assessment of intraoperative surgical margins by examining a surgical specimen oriented through a 3D-printed specific patient tongue with real-time Magnetic Resonance Imaging (MRI). We reported the preliminary results of a case series of 10 patients, prospectively enrolled, with oral tongue carcinoma who underwent surgery between February 2020 and April 2021. Two radiologists with 5 and 10 years of experience, respectively, in Head and Neck radiology in consensus evaluated specimen MRI and measured the distance between the tumor and the specimen surface. We performed intraoperative bedside FSA. To compare the performance of bedside FSA and MRI in predicting definitive margin status we computed the weighted sensitivity (SE), specificity (SP), accuracy (ACC), area under the ROC curve (AUC), F1-score, Positive Predictive Value (PPV), and Negative Predictive Value (NPV). To express the concordance between FSA and ex-vivo MRI we reported the jaccard index. Results: Intraoperative bedside FSA showed SE of 90%, SP of 100%, F1 of 95%, ACC of 0.9%, PPV of 100%, NPV (not a number), and jaccard of 90%, and ex-vivo MRI showed SE of 100%, SP of 100%, F1 of 100%, ACC of 100%, PPV of 100%, NPV of 100%, and jaccard of 100%. These results needed to be validated in a larger sample size of 21- 44 patients. Conclusion: The presented method allows a more accurate evaluation of surgical margin status, and the first clinical experiences underline the high potential of integrating FSA with ex-vivo MRI of the fresh surgical specimen.

restrictedJan 2023View details →

Dataset related to the article "A token-mixer architecture for CAD-RADS classification of coronary stenosis on multiplanar reconstruction CT images"

This record contains raw data related to the article "A token-mixer architecture for CAD-RADS classification of coronary stenosis on multiplanar reconstruction CT images" A B S T R A C T Background and objective: In patients with suspected Coronary Artery Disease (CAD), the severity of stenosis needs to be assessed for precise clinical management. An automatic deep learning-based algorithm to classify coronary stenosis lesions according to the Coronary Artery Disease Reporting and Data System (CAD-RADS) in multiplanar reconstruction images acquired with Coronary Computed Tomography Angiography (CCTA) is proposed. Methods: In this retrospective study, 288 patients with suspected CAD who underwent CCTA scans were included. To model long-range semantic information, which is needed to identify and classify stenosis with challenging appearance, we adopted a token-mixer architecture (ConvMixer), which can learn structural relationship over the whole coronary artery. ConvMixer consists of a patch embedding layer followed by repeated convolutional blocks to enable the algorithm to learn long-range dependences between pixels. To visually assess ConvMixer performance, Gradient-Weighted Class Activation Mapping (Grad-CAM) analysis was used. Results: Experimental results using 5-fold cross-validation showed that our ConvMixer can classify significant coronary artery stenosis (i.e., stenosis with luminal narrowing ≥50%) with accuracy and sensitivity of 87% and 90%, respectively. For CAD-RADS 0 vs. 1–2 vs. 3–4 vs. 5 classification, ConvMixer achieved accuracy and sensitivity of 72% and 75%, respectively. Additional experiments showed that ConvMixer achieved a better trade-off between performance and complexity compared to pyramid-shaped convolutional neural networks. Conclusions: Our algorithm might provide clinicians

restrictedJul 2023View details →

Dataset related to the article "A deep-learning approach for myocardial fibrosis detection in early contrast-enhanced cardiac CT images"

This record contains raw data related to the article "A deep-learning approach for myocardial fibrosis detection in early contrast-enhanced cardiac CT images" Aims: Diagnosis of myocardial fibrosis is commonly performed with late gadolinium contrast-enhanced (CE) cardiac magnetic resonance (CMR), which might be contraindicated or unavailable. Coronary computed tomography (CCT) is emerging as an alternative to CMR. We sought to evaluate whether a deep learning (DL) model could allow identification of myocardial fibrosis from routine early CE-CCT images. Methods and results: Fifty consecutive patients with known left ventricular (LV) dysfunction (LVD) underwent both CE-CMR and (early and late) CE-CCT. According to the CE-CMR patterns, patients were classified as ischemic (n =&thinsp;15, 30%) or non-ischemic (n =&thinsp;35, 70%) LVD. Delayed enhancement regions were manually traced on late CE-CCT using CE-CMR as reference. On early CE-CCT images, the myocardial sectors were extracted according to AHA 16-segment model and labeled as with scar or not, based on the late CE-CCT manual tracing. A DL model was developed to classify each segment. A total of 44,187 LV segments were analyzed, resulting in accuracy of 71% and area under the ROC curve of 76% (95% CI: 72%−81%), while, with the bull’s eye segmental comparison of CE-CMR and respective early CE-CCT findings, an 89% agreement was achieved. Conclusions: DL on early CE-CCT acquisition may allow detection of LV sectors affected with myocardial fibrosis, thus without additional contrast-agent administration or radiational dose. Such tool might reduce the user interaction and visual inspection with benefit in both efforts and time.

CCDI Data Catalog8/100

A Normative Dataset of Healthy Pediatric Cranial Computed Tomography (CT) Images

This dataset contains cranial CT images from 100 healthy pediatric subjects, aged one month to ten years, to address the limited availability of public normative reference data for healthy pediatric cranial CT imaging. Supporting data includes subject demographics (age, sex, race/ethnicity) and a Python script for ease of data loading. It is designed to support research reproducibility by providing a common baseline for developing and validating AI algorithms, conducting normative studies of neurodevelopment, and serving as a control cohort for pediatric neurological research.

unknownView details →

restrictedMay 2024View details →

ZooScan77: ZooScan zooplankton image dataset for Migratory Crossroads ML/DL model training and evaluation

<h2>ZooScan77</h2> This dataset contains a collection of ca. 1,000,000 images of 77 zooplankton, micronekton and detritus categories captured by ZooScan imager. The dataset is not public and must only be used within the scope of Migratory Crossroads project only.

restrictedAug 2019View details →

dataset for image analysis for ML

image dataset for image analysis for ML

restrictedJun 2023View details →

Dataset of "Parts-per-Object Count in Agricultural Images: Solving Phenotyping Problems via a Single Deep Neural Network" paper

This includes the relevant datasets to:  Khoroshevsky, F., Khoroshevsky, S., & Bar-Hillel, A. (2021). Parts-per-object count in agricultural images: Solving phenotyping problems via a single deep neural network. Remote Sens. 13(13), 2496. https://doi.org/10.3390/rs13132496   Datasets related to wheat and banana are not public since it belongs to the Israel Phenomics consortium. This research was funded by the Generic technological R&D program of the Israel innovation authority-the Phenomics consortium, and the Ministry of Science & Technology, Israel.

restrictedMay 2021View details →

An image dataset for estimation of nuclei density of lobules related to Cell atlas of the regenerating human liver after portal vein embolization

Dataset of Tissue samples for cell atlas datafilename coding: first Character: sex of patient ( B- men, K- women)  second char: condition (0- conroll, 1- embolized, 2- regenerated) next chars to Letter: Date of Dataset Letter char with number: Lobule of Marker Charge( G- GLUL) Tiff: fiji proceed images zip: coordinates for fiji csv: cleaned and readable format of coordinates for python PNG: Hoechst marked Nuclei of GLUL marker

restrictedJul 2020View details →

SK inventory [dataset] misc. images version 1.0 28/07/2020

SK inventory [dataset] misc. images version 1.0 28/07/2020

restrictedNov 2015View details →

Image Dataset

Dataset de prueba para imágenes

restrictedJan 2023View details →

The EMory BrEast imaging Dataset (EMBED): A Racially Diverse, Granular Dataset of 3.4 Million Screening and Diagnostic Mammographic Images

Open the record for dataset details and reuse information.