Skip to content
Open Imaging Index
MRI Prostate

PROSTATE-3T

Prostate-3T (Radboud University Nijmegen Medical Centre)

Prostate MRI taken for cancer detection, released for a 2013 challenge on segmenting prostate structures, with central gland and peripheral zone masks for some cases. 64 men from a Dutch hospital with axial T2-weighted 3T scans, 30 with masks.

Go to dataset Edit Report

Overview

PROSTATE-3T is a small collection of prostate MRI from Radboud University Nijmegen Medical Centre in the Netherlands, contributed by Geert Litjens, Jurgen Futterer and Henkjan Huisman and published by The Cancer Imaging Archive (TCIA) in 2013. The scans were taken for prostate cancer detection and were released as source images for the NCI-ISBI 2013 challenge on automated segmentation of prostate structures. It is mainly used for prostate gland and zonal segmentation.

Composition

The collection snapshot lists 64 men. The public TCIA metadata service returns 64 patients, each with one study and one axial T2-weighted series, for 1,258 DICOM images in total. Ages recorded in the DICOM headers range from 51 to 78 years (median 64). The download is about 270 MB.

Acquisition

All series are transversal turbo spin echo T2-weighted acquisitions at 3 T. The collection page names a Siemens TrioTim with a pelvic phased-array coil and no endorectal coil. Per the series metadata, 50 series come from a TrioTim and 14 from a Siemens Skyra. In the NCI-ISBI challenge description, slices from this site are 4 mm thick.

Annotations

Two segmentation sets are open downloads on the collection page, both built with MeVisLab. The first holds central gland and peripheral zone masks as NRRD files for 30 cases, the training cases of the NCI-ISBI 2013 challenge. The second holds seminal vesicle and neurovascular bundle masks as MHA files for 15 other cases, prepared for a follow-up challenge that never took place. The challenge's leaderboard and test masks, which mix this collection with PROSTATE-DIAGNOSIS, are distributed separately on the challenge page.

Known limitations

  • The Data Access table on the collection page lists 59 subjects and 1,161 images, while the collection snapshot and the metadata service give 64 subjects and 1,258 images.
  • Three series have protocol names that mention an endorectal coil (ERC), although the collection summary states a pelvic phased-array coil only.
  • No clinical data, pathology, cancer grade or lesion annotations are included.
  • Only one sequence is available per patient, with no diffusion or perfusion imaging.

Cohort

Aggregate numbers from the sources below. Bars are relative to the 64 subjects.

Sex

  • Male 64 100%

PatientSex in TCIA public NBIA API, getPatientStudy for collection Prostate-3T (CC BY 3.0, no login)

Modality

  • MRI 64 100%

Modality in TCIA public NBIA API, getSeries for collection Prostate-3T (CC BY 3.0, no login)

Scanner vendor

  • Siemens Healthineers 64 100%

Manufacturer in TCIA public NBIA API, getSeries for collection Prostate-3T (CC BY 3.0, no login)

Country

  • Netherlands 64 100%

Summary in TCIA analysis result page NCI-ISBI 2013 Challenge, Automated Segmentation of Prostate Structures

Age

mean 63.5 ± 6.3, median 64, range 51 to 78

0
10
20
30
37
50-5960-6970-79

PatientAge in TCIA public NBIA API, getPatientStudy for collection Prostate-3T (CC BY 3.0, no login)

Contrast / sequence

Groups can overlap

  • T2-weighted 64 100%

ProtocolName in TCIA public NBIA API, getSeries for collection Prostate-3T (CC BY 3.0, no login)

Field strength

  • 3 T 64 100%

Summary in TCIA Prostate-3T collection page (version 1, updated 2013/05/31)

Contrast combinations

How many subjects have exactly each set of contrasts.

T2wSubjects with exactly this set
64

ProtocolName in TCIA public NBIA API, getSeries for collection Prostate-3T (CC BY 3.0, no login)

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

Access
Open download

Download without an account

Images download with the TCIA Data Retriever; the two segmentation archives are direct ZIP downloads from the collection page.

Access 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.

Original license text Version read: 3.0 Checked 2026-10-07

What you can do

  • Yes
  • 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

Litjens G, Futterer J, Huisman H. (2015). Data From Prostate-3T. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2015.QJTV5IL5

Sources

Every number on this page comes from one of these documents. Each chart names the table or page it is taken from. The raw numbers are in stats.csv.