PROSTATE-DIAGNOSIS
1.5 T endorectal-coil prostate MRI of 92 prostate cancer patients with T2-weighted, T1-weighted and dynamic contrast-enhanced series, plus pathology and radiology reports and small sets of manual zonal and structure segmentations.
Overview
PROSTATE-DIAGNOSIS is a collection of diagnostic prostate MRI from patients with prostate cancer, contributed by B. Nicolas Bloch, Ashali Jain and C. Carl Jaffe and hosted by The Cancer Imaging Archive. Images were released in January 2013 and received a DOI in 2015. Besides the DICOM images, the collection offers free-text pathology and radiology reports and three small sets of manual segmentations, one of which served as training data for the NCI-ISBI 2013 prostate segmentation challenge. It is used for prostate gland and zonal segmentation and for work on lesion detection with multiparametric MRI.
Composition
The collection snapshot lists 92 subjects and 5.67 GB of MR images. The public series metadata hold one study per patient with 368 series and 32,537 images. Every patient has axial and coronal T2-weighted turbo spin echo series and a dynamic contrast-enhanced series. Eighty patients also have an axial T1-weighted series without contrast, and seven have a contrast-enhanced axial T1-weighted series instead. A clinical spreadsheet covers 54 subjects with biopsy and prostatectomy pathology reports and the MRI radiology report, as unstructured text.
Acquisition
All scans come from a 1.5 T Philips Achieva with a combined surface and endorectal coil. Dynamic contrast-enhanced images were taken before, during and after intravenous injection of 0.1 mmol/kg gadolinium-DTPA. Study dates range from 2005 to 2010.
Annotations
- Multi-component 3D Slicer label maps for 5 cases on the axial T2-weighted series, outlining the gland, central gland, peripheral zone, seminal vesicles, urethra, dominant cancer nodule, neurovascular bundles, penile bulb, ejaculatory duct, verumontanum and rectum.
- Central gland and peripheral zone label maps for 30 subjects from the NCI-ISBI 2013 challenge, in NRRD.
- Seminal vesicle and neurovascular bundle masks for 15 subjects in MHA, prepared for a follow-up challenge that did not take place.
Known limitations
- The Data Access table (87 subjects, 87 studies, 348 series, 30,903 images) disagrees with the collection snapshot and the series metadata (92 subjects, 368 series).
- No demographics or structured clinical variables; the reports are free text and cover only part of the cohort.
- Single scanner at a single field strength, without diffusion-weighted imaging.
- Manual segmentations exist for at most 30 subjects.
Cohort
Aggregate numbers from the sources below. Bars are relative to the 92 subjects.
Contrast combinations
How many subjects have exactly each set of contrasts.
| T1w | T1w_ce | T2w | dce | Subjects with exactly this set |
|---|---|---|---|---|
80 | ||||
7 | ||||
5 |
Modality
subjects
- MRI 92 100%
Contrast / sequence
subjects, values can overlap
- T2-weighted 92 100%
- DCE perfusion 92 100%
- T1-weighted 80 87%
- T1-weighted, contrast enhanced 7 8%
Scanner vendor
subjects
- Philips 92 100%
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
DICOM images via the TCIA Data Retriever; clinical spreadsheet and segmentation archives as direct downloads from the collection page.
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.
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
Bloch BN, Jain A, Jaffe CC. (2015). Data From PROSTATE-DIAGNOSIS [Dataset]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2015.FOQEUJVT
All numbers
Every number on this page, as stored in stats.csv, with its source.
| Measure | Breakdown | Value | Source |
|---|---|---|---|
| Subjects | total The Data Access table lists 87 subjects, 87 studies, 348 series and 30,903 images | 92 | tcia-prostate-diagnosis Collection Snapshot |
| Studies | total one MRI study per patient | 92 | tcia-prostate-diagnosis-series StudyInstanceUID |
| Scans | total DICOM series | 368 | tcia-prostate-diagnosis-series SeriesInstanceUID |
| Images | total DICOM images (slices) summed over series | 32,537 | tcia-prostate-diagnosis-series ImageCount |
| Subjects | modality=MR | 92 | tcia-prostate-diagnosis-series PatientID, Modality |
| Subjects | vendor=philips all series from a Philips Achieva | 92 | tcia-prostate-diagnosis-series PatientID, Manufacturer |
| Subjects | contrast=T2w T2W_TSE axial and coronal series | 92 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast=dce AX BLISS_GAD, AX BLISS_PRE and dyn_THRIVE series | 92 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast=T1w T1W_TSE_AX and T1W_TSE_AX_NO GAD series | 80 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast=T1w_ce T1W_TSE_AX_Gd series | 7 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast_set=dce+T1w+T2w | 80 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast_set=dce+T1w_ce+T2w | 7 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
| Subjects | contrast_set=dce+T2w | 5 | tcia-prostate-diagnosis-series PatientID, StudyInstanceUID, SeriesDescription |
Sources
The keys used in the table above.
- tcia-prostate-diagnosis TCIA PROSTATE-DIAGNOSIS collection page (version 2, updated 2021/08/09) website
- tcia-prostate-diagnosis-series TCIA NBIA API series metadata for PROSTATE-DIAGNOSIS (CC BY 3.0, open, no account) computed