Off-Grid Ultrasound Imaging by Stochastic Optimization

This is the code repository accompanying the paper Off-Grid Ultrasound Imaging by Stochastic Optimization. INverse grid-Free Estimation of Reflectivities (INFER) is a method for off-grid ultrasound imaging. Instead of beamforming RF data, we construct a matrix-free forward model of the acquisition process and optimize for the reflectivities, backscatter source locations, and other tissue and transducer parameters directly.

Installing the dependencies

The code was tested with python 3.10. The dependencies can be installed by running

pip install -r requirements.txt

After that you also need to install JAX. To install JAX, please refer to the official installation guide. At the time of writing the command to install JAX is

pip install -U "jax[cuda12]"

Downloading the data

The data can be downloaded from the Zenodo repository. The files should be placed in the data folder.

Running the code

Running an example

The script run_example.py runs the optimization on the phantom data.

python3 run_example.py

Generating the results from the paper

To reproduce the plots from the paper, run the script run_generate_paper_results.py. This will run all the experiments with INFER and the baselines. The generated figures will be placed in the results/figures folder. Be aware that this will take some time.

python3 run_generate_paper_results.py