OA-Breast Database

Optical and Acoustic Breast Phantom Database (OA-Breast)

Download link: OA-BreastDownload
We STRONGLY recommend joining our mailing list to keep updated with the latest changes of the dataset!

This database includes a collection of numerical breast phantoms for use in various optical and acoustic imaging simulation studies, including photoacoustic imaging, ultrasound computed tomography, diffuse optical tomography, etc. The numerical phantoms are generated from clinical Magnetic Resonance Angiography (MRA) data collected from Washington University in St. Louis, School of Medicine with the approval of IRB.

  1. Two data-types are available for download:
    1. You can directly download the binary data from each separate folder “./Neg_$(ID)_Left”, each dataset contains the following phantom files:

    1. MergedPhantom.DAT: the generated breast phantom file, stored in binary data format in uint8 (in Unix Matlab R2015a).
    2. PreData_Interp.DAT, PostData_Interp.DAT: the original MRI breast data files (after interpolated onto a uniform grid), stored in binary data format in float32 (in Unix Matlab R2015a).
      If you are limited by download speed, you can choose to download the merged phantom file only if you don’t need the original MR data. The original MRI data files are pretty large.
  2. We also provide the same phantom files in the HDF5 format, contained in the folder “./hdf5/”.

Each dataset represents one breast of a healthy female (left or right), and includes the following items (the 3D rendered view of one dataset is given on the bottom):

  1. The merged breast phantom MergedPhantom.DAT: a binary file written under UINT8 data-format in Linux OS. It can be reshaped into a 3D matrix, where each voxel’s value represents one specific tissue type: background is 0, fibroglandular tissue is 2, fat is 3, skin is 4, and vessel is 5. Corresponding speed of sound, density, optical absorption, and optical scattering phantoms can be easily created by assigning values to different tissue types within this phantom.
  2. The pre-contrast MRI dataset and the post-contrast MRI dataset PreData_Interp.DAT, PostData_Interp.DAT, which correspond to the original MRI data used to derive the specific breast phantom.
  3. A txt file specifying the details of the phantom README.txt. 
  4. An example MATLAB script that loads in the phantom files SCRP_ReadDataExample.m.

If you are limited by downloading speed or storage and only wanted to use the phantoms, you can skip the two MRI datasets, since we only include them in case someone is interested in the original MR data.

We welcome anyone to use our database for education and research purpose. The database is free to download and no registration is required. However, we would greatly appreciate it if you can send us an email (louy AT including your name, institute, and the research project you want to use the database for, so that we can further improve the database. If you want to cite our database, please cite the following paper:

Yang Lou, Weimin Zhou, Thomas P. Matthews, Catherine M. Appleton, Mark A. Anastasio, “Generation of anatomically realistic numerical phantoms for photoacoustic and ultrasonic breast imaging,” J. Biomed. Opt. 22(4), 041015 (2017), doi: 10.1117/1.JBO.22.4.041015. 

If you have any questions, comments, or suggestions, please contact Yang Lou at, and we will get back to you as soon as possible.

This OA-Breast Database is made available under the Open Database License: Any rights in individual contents of the database are licensed under the Database Contents License:



  • 01. 31. 2017: Updated the description of the database. (Yang Lou)
  • 01. 25. 2017: Pushed the most recent results to the repository, all phantoms are now consistent with the journal paper.  (Yang Lou)
  • 03. 20. 2017: Restructured the README file.
  • 06. 14. 2017: fixed a wrong dimension number in the home folder README file (thanks to Dr. Bernhard Kaplan for the catch-up!)
  • 09. 22. 2017: added hdf5-type data for download, for portability across platforms (thanks to Dr. Liren Zhu from CalTech for his help! )