This repository contains an image-based method for cardio-respiratory phase estimation, gating, and temporal super-resolution of cardiac ultrasound.
The core algorithm is implemented in the class uspgs.USPGS. Please see the docstrings to understand its functionality.
There are two jupyter notebooks that will further help understand how to use the method:
- publication_experiments.ipynb - contains the code used to generate the results/figures for a paper.
- WinProbe-Experiments.ipynb - shows how to apply the algorithm on WinProbe data.
The following python packages need to be installed in your virtual environment to be able to run the code:
- numpy
- scipy
- scikit-learn
- statsmodels
- matplotlib
- jupyter
- angles
- SimpleITK
- MedPy
- opencv-python
You can run the following command to install these dependencies:
pip install -U -r requirements.txt
The code uses opencv's Python interface for reading, displaying, and writing videos.
pip installing the requirements using the aforementioned command attempts to
install opencv's python interface by installing the opencv-python>=3.2.0.7
package from PyPI for convenience.
However since this is not an official release of opencv, it may cause issues, especially on OSX and Linux platforms. If this happens, download and install opencv binaries for your operating system from the official opencv website. Follow the instructions here to install OpenCV-Python on your operating system.