CT Scan Datasets — DICOM Computed Tomography Data
CT scan datasets provide cross-sectional, three-dimensional imaging reconstructed from multiple X-ray projections, offering far greater anatomical detail than plain radiography. Computed tomography is indispensable for training models in oncology, trauma, neuroimaging, and pulmonary disease because it resolves soft tissue, bone, and vasculature in volumetric detail. A CT dataset is typically delivered as DICOM image series, axial slices that reconstruct into isotropic volumes, annotated with slice thickness, reconstruction kernel, contrast phase, kVp, and mAs in the metadata.
Studies span non-contrast and contrast-enhanced acquisitions, multiphase protocols (arterial, venous, delayed), and specialized techniques such as CT angiography (CTA), low-dose lung-cancer screening, and dual-energy CT. Clinically valuable CT datasets cover a wide range of regions and findings: intracranial hemorrhage and ischemic stroke on head CT; pulmonary nodules, emphysema, and interstitial lung disease on chest CT; liver, kidney, and pancreatic lesions on abdominal CT; pulmonary embolism on CTA; and fractures and internal injuries in trauma protocols. The most useful datasets include voxel-level segmentation masks of organs, lesions, and abnormalities, along with radiologist-confirmed labels, lesion measurements following RECIST, and Hounsfield-unit calibration.
High-quality cohorts document acquisition parameters, balance pathology prevalence, and span multiple scanner vendors and reconstruction settings so models generalize beyond a single site. Rigorous de-identification removes PHI from DICOM headers and defaces or skull-strips head CT volumes where required, while preserving diagnostic fidelity. On GetDATA, clients post CT requests specifying body region, contrast phase, slice thickness, annotation type (volumetric segmentation, bounding box, or study-level label), label taxonomy, and minimum case counts, and verified providers fulfill them with compliant, quality-scored CT data in DICOM.
Beyond diagnosis, CT datasets power radiomics pipelines that extract quantitative texture and shape features, support automated organ-at-risk contouring for radiotherapy planning, and enable opportunistic screening for osteoporosis, coronary calcium, and body composition from scans acquired for unrelated indications. Browse the open CT scan requests below, or explore related imaging categories.