Skip to content
Open Imaging Index
CT Mammography MRI PET Breast

BREAST-DIAGNOSIS

Diagnostic breast MRI of women with breast cancer, DCIS, benign findings or high risk, with mammography for some and PET/CT for a few, plus clinical, pathology and BI-RADS data. 88 women scanned at 1.5 T with T2, STIR and dynamic contrast-enhanced series.

Go to dataset Edit Report

Overview

BREAST-DIAGNOSIS is a collection of diagnostic breast imaging from 88 women, contributed by B. Nicolas Bloch, Ashali Jain and C. Carl Jaffe and hosted by The Cancer Imaging Archive. The cases cover high-risk patients without cancer, ductal carcinoma in situ, fibroadenomas and invasive lobular and ductal carcinomas. Images were released in November 2011 and received a DOI in 2015. A clinical spreadsheet adds BI-RADS findings from the MRI report and abstracts of the pathology report.

Composition

The collection holds 148 studies, 429 series and 105,050 DICOM images (60.87 GB). Breast MRI covers 84 subjects in 112 studies; 24 of them have more than one MRI time point. Digital mammograms or radiographs of surgical specimens exist for 17 subjects, and 3 subjects have whole-body PET/CT studies. The spreadsheet lists 51 cases, 50 of which have images, and gives the affected side, a free-text background, the pathology diagnosis, ER, PR, HER2, Ki67 and E-cadherin results, Oncotype DX scores where available, the age decade, the BI-RADS category and the report impression.

Acquisition

MRI was acquired on Philips Achieva and Intera scanners at 1.5 T. Most MRI studies contain a T2-weighted turbo spin echo series, a STIR series and BLISS dynamic contrast-enhanced series; a few add T1-weighted, diffusion tensor or ADC series. Magnevist was injected over 6 to 7 seconds at 3 ml/s, with a dose in millilitres equal to 10% of body weight in pounds, and the first dynamic series started one minute after injection began. Mammography came from Lorad (Hologic) units and PET/CT from a GE Discovery STE. Study dates range from 2006 to 2009.

Annotations

The spreadsheet notes the BI-RADS category and pathology diagnosis per case. There are no segmentations.

Known limitations

  • Clinical and pathology entries are free text with inconsistent wording and cover only part of the cohort.
  • The age decade is given for 19 of the imaged subjects; no other demographics are provided.
  • The DICOM patient weight tag is in kilograms, while the contrast dose rule uses pounds.
  • One MRI field strength; mammography and PET/CT exist for few subjects.

Cohort

Aggregate numbers from the sources below. Bars are relative to the 88 subjects.

Sex

  • Female 88 100%

PatientSex in TCIA NBIA API patient metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

Modality

  • MRI 84 95%
  • Mammography 17 19%
  • CT 3 3%
  • PET 3 3%

PatientID, Modality in TCIA NBIA API series metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

Scanner vendor

  • Philips 84 95%
  • Hologic 17 19%
  • GE HealthCare 3 3%

PatientID, Manufacturer in TCIA NBIA API series metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

Age

0
2
4
6
8
8
20-2940-4950-5960-6970-79

Age decade in BREAST-DIAGNOSIS clinical spreadsheet, updated 7/16/11 (CC BY 3.0, open, no account)

Contrast / sequence

Groups can overlap

  • T2-weighted 83 94%
  • STIR 83 94%
  • DCE perfusion 83 94%
  • T1-weighted 3 3%
  • Diffusion-weighted 1 1%
  • ADC map 1 1%

PatientID, Modality, SeriesDescription in TCIA NBIA API series metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

View

  • Craniocaudal 12 14%
  • Mediolateral oblique 5 6%

PatientID, SeriesDescription in TCIA NBIA API series metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

Contrast combinations

How many subjects have exactly each set of contrasts.

T1wT2wSTIRdwiADCdceSubjects with exactly this set
76
3
1
1
1
1
1

PatientID, Modality, SeriesDescription in TCIA NBIA API series metadata for BREAST-DIAGNOSIS (CC BY 3.0, open, no account)

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

DICOM images via the TCIA Data Retriever; the clinical spreadsheet as a direct download 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

Bloch BN, Jain A, Jaffe CC. (2015). BREAST-DIAGNOSIS [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2015.SDNRQXXR

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.