Longitudinal multi-parametric MRI + 18F-FDG PET/CT collection of locally
advanced cervical cancer patients undergoing chemoradiation, with
radiation-oncologist tumor contours (RTSTRUCT) defined on T2-weighted MRI as
the gold standard. Each patient was imaged at three time points spanning the
chemoradiation course.
Dataset Details
Field
Value
Modalities
MR (T1W, T2W, DCE, DWI), 18F-FDG PET, CT
Body part
Cervix / pelvis
Task
3D tumor segmentation on T2W MRI (longitudinal)
Patients
23 (CCTH-A01 âĻ CCTH-A12, CCTH-B01 âĻ CCTH-B11)
Time points
3 per patient â Baseline (0 Gy) / Early (20â25 Gy, 2â2.5 wk) / Mid-treatment (45â50 Gy, 4â5 wk)
Studies
171
Series
821
Images
131,556 DICOM slices
MR series
523 (15.91 GB)
CT series
65 (11.48 GB)
PET series
68 (0.99 GB)
RTSTRUCT files
68 (â 23 patients à 3 time points; CCTH-A06 has 2)
Tumor contours are stored as DICOM RTSTRUCT files in segmentations/.
Each RTSTRUCT references a single MR T2W sagittal series via
ReferencedFrameOfReferenceSequence â RTReferencedStudySequence â RTReferencedSeriesSequence.SeriesInstanceUID â that referenced series is the
image volume the contour is defined on. ROI names follow Ut-MRT2-Sag-{1|2|3}
where the trailing index matches the time point (1 = Baseline, 2 = Early,
3 = Mid-treatment), and StructureSetName carries Timepoint{1|2|3}.
The collection paper describes contours as tumor volume only â no OARs and
no lymph nodes, drawn by the study radiation oncologists with T2W MRI
chosen for its highest soft-tissue contrast for cervical tumor delineation.
Structure
text
images_mr////*.dcm # MR (T1W, T2W, DCE, DWI variants)
images_ct////*.dcm # CT
images_pt////*.dcm # PET
segmentations////*.dcm # RTSTRUCT (one per time point)
registrations////*.dcm # DICOM REG (PET/CT â MR coregistration)
series_to_patient.json # series-level index
PatientID ranges over CCTH-A01..A12 and CCTH-B01..B11. StudyDescription
in each RTSTRUCT carries MR1 / MR2 / MR3, identifying the time point.
CCTH-B11 has no CT series; CCTH-A06 has 2 RTSTRUCTs (one time point missing).
Released: 2023-01-20 (now fully public, no registration required)
Citation
bibtex
@misc{mayr2023ccth,
author = {Mayr, N. and Yuh, W. T. C. and Bowen, S. and Harkenrider, M. and
Knopp, M. V. and Lee, E. Y.-P. and Leung, E. and Lo, S. S. and
Small Jr., W. and Wolfson, A. H.},
title = {Cervical Cancer -- Tumor Heterogeneity: Serial Functional and
Molecular Imaging Across the Radiation Therapy Course in
Advanced Cervical Cancer (Version 1) [Data set]},
year = {2023},
publisher = {The Cancer Imaging Archive},
doi = {10.7937/ERZ5-QZ59}
}
@article{bowen2017radiomic,
author = {Bowen, S. R. and Yuh, W. T. C. and Hippe, D. S. and Wu, W. and
Partridge, S. C. and Elias, S. and Jia, G. and Huang, Z. and
Sandison, G. A. and Nelson, D. and Knopp, M. V. and Lo, S. S. and
Kinahan, P. E. and Mayr, N. A.},
title = {Tumor radiomic heterogeneity: Multiparametric functional imaging
to characterize variability and predict response following
cervical cancer radiation therapy},
journal = {Journal of Magnetic Resonance Imaging},
volume = {47},
number = {5},
pages = {1388--1396},
year = {2017},
doi = {10.1002/jmri.25874}
}