Breast Cancer-Related Lymphedema (BCRL) Database

The repository includes the code and dataset associated with the following publication:
Sobhan Goudarzi, Jesse Whyte, Mathieu Boily, Anna Towers, Robert D. Kilgour, and Hassan Rivaz, “Segmentation of Arm Ultrasound Images in Breast Cancer-Related Lymphedema: A Database and Deep Learning Algorithm”, IEEE Transactions on Biomedical Engineering (TBME), 2023, in press.

The dataset includes 468 ultrasound images collected from two groups of women. The patient group includes 19 women with Stage 2 BCRL, and the control group contains 20 healthy women. As the above figure illustrates, ultrasound images were collected from six different locations on each arm of the subjects.

The codes were written in Python and using the PyTorch library.

Instructions for usage:

  • All images are provided with “.mat” format.
  • There exists a “readme.pdf”  file containing details that you may need for your application.
  • The implementation code of the proposed network is provided in python (using PyTorch library).
  • There exists a flow chart that explains how the main Pytorch script works, and each file, in the Code folder, talks to one another.