PROSTATE-MRI
Endorectal-coil 3T multiparametric prostate MRI (T2-weighted, diffusion, DCE) of 26 men with biopsy-proven prostate cancer before robotic radical prostatectomy, with whole-mount histopathology cut in the MRI plane using a patient-specific mold.
Overview
PROSTATE-MRI is a small radiology-pathology collection from the National Cancer Institute in Bethesda, Maryland, published on The Cancer Imaging Archive (TCIA). It pairs preoperative prostate MRI with whole-mount histopathology of the removed gland from 26 men with biopsy-confirmed prostate cancer who then had robotic-assisted radical prostatectomy. The data were acquired between 2008 and 2010. Because the specimen was sectioned in the same plane as the MRI, the collection suits work on registering histology to MRI and on mapping tumor appearance in multiparametric MRI to pathology.
Composition
The collection holds 26 subjects with one MR study each, 182 DICOM series and 22,036 MR images, plus 26 whole-mount histopathology images in JPEG, one per subject. According to the open series metadata, every subject has axial, coronal and sagittal T2-weighted turbo spin echo series, two diffusion-weighted series and a dynamic contrast-enhanced series with a pre-contrast acquisition. No clinical data such as age, PSA or Gleason score are distributed.
Acquisition
All MRI was acquired at 3T on a Philips Achieva with an endorectal coil combined with a phased-array surface coil. The investigators state that the diffusion b values were 0, 188, 375, 563 and 750 s/mm², and that these values may no longer be present in the DICOM headers. After surgery, a mold built from each patient's MRI held the specimen, which was then cut in the MRI plane.
Annotations
The collection does not include tumor or gland segmentations. The histopathology images serve as the reference for tumor location.
Known limitations
- 26 subjects from one center and one scanner model.
- Only one whole-mount histopathology image per subject is distributed.
- The DWI b values must be taken from the collection page, not from the DICOM headers.
- Downloads were restricted to users approved by the investigators until May 2018.
Cohort
Aggregate numbers from the sources below. Bars are relative to the 26 subjects.
Sex
- Male 26 100%
Contrast combinations
How many subjects have exactly each set of contrasts.
| T2w | dwi | dce | Subjects with exactly this set |
|---|---|---|---|
26 |
Modality
subjects
- MRI 26 100%
- Pathology slides 26 100%
Contrast / sequence
subjects, values can overlap
- T2-weighted 26 100%
- Diffusion-weighted 26 100%
- DCE perfusion 26 100%
Scanner vendor
subjects
- Philips 26 100%
Field strength
subjects
- 3 T 26 100%
Country
subjects
- United States 26 100%
License and access
Our reading of the license, not legal advice. Before you use the data, read the original license and confirm that your use is allowed. We take no responsibility for how you use a dataset. Full disclaimer
Download without an account
MR images download with the TCIA Data Retriever, histopathology JPEGs through IBM Aspera. Access was restricted to users approved by the investigators until May 2018 and is open since then.
Creative Commons Attribution 3.0 Unported
Older version of CC BY. Use, share and adapt the data for any purpose, including commercial use, as long as you credit the creators.
TCIA lists both the MR images and the histopathology images under CC BY 3.0. Users must also follow the TCIA Data Usage Policy, which says users must "Not use the requested datasets, either alone or in concert with any other information, to identify or contact individual participants" and must cite the dataset DOI (https://www.cancerimagingarchive.net/data-usage-policies-and-restrictions/).
What you can do
- Yes
- Yes
- Yes
- Yes
What you can share
- Yes
- Yes
- Yes
What you must do
- Yes
- Share alike No
- No
- No
- Manuscript review No
- Release code No
- Return results No
- Delete after use No
Limits
- No
- Location limits No
Citation
Choyke P, Turkbey B, Pinto P, Merino M, Wood B. (2016). Data From PROSTATE-MRI. The Cancer Imaging Archive. http://doi.org/10.7937/K9/TCIA.2016.6046GUDv
All numbers
Every number on this page, as stored in stats.csv, with its source.
| Measure | Breakdown | Value | Source |
|---|---|---|---|
| Subjects | total | 26 | tcia-prostate-mri Data Access table |
| Studies | total MR studies | 26 | tcia-prostate-mri Data Access table |
| Scans | total sum of the disjoint MR (182) and histopathology (26) series rows | 208 | tcia-prostate-mri Data Access table |
| Images | total sum of the disjoint MR (22,036) and histopathology (26) image rows | 22,062 | tcia-prostate-mri Data Access table |
| Scans | modality=MR DICOM series | 182 | tcia-prostate-mri Data Access table |
| Images | modality=MR DICOM images (slices) | 22,036 | tcia-prostate-mri Data Access table |
| Scans | modality=SM histopathology series | 26 | tcia-prostate-mri Data Access table |
| Images | modality=SM whole-mount histopathology JPEGs, one per subject | 26 | tcia-prostate-mri Data Access table |
| Subjects | modality=MR | 26 | tcia-prostate-mri Data Access table |
| Subjects | modality=SM | 26 | tcia-prostate-mri Data Access table |
| Subjects | sex=male | 26 | prostate-mri-nbia-digest Patient Sex |
| Subjects | field_strength=3 | 26 | tcia-prostate-mri |
| Subjects | vendor=philips Philips Achieva | 26 | prostate-mri-nbia-digest Manufacturer |
| Subjects | country=US National Cancer Institute, Bethesda | 26 | tcia-prostate-mri |
| Subjects | contrast=T2w T2 TSE axial, coronal and sagittal series | 26 | prostate-mri-nbia-digest Series Description |
| Subjects | contrast=dwi two DWI series per subject | 26 | prostate-mri-nbia-digest Series Description |
| Subjects | contrast=dce dynamic series plus a pre-contrast series | 26 | prostate-mri-nbia-digest Series Description |
| Subjects | contrast_set=dce+dwi+T2w | 26 | prostate-mri-nbia-digest Series Description |
Sources
The keys used in the table above.
- tcia-prostate-mri TCIA PROSTATE-MRI collection page (version 1, updated 2011/06/30) website
- prostate-mri-nbia-digest NBIA metadata digest of the MR series (open download, CC BY 3.0), counted per Patient ID and Series Description computed