Freehand ultrasound without external trackers

This repository contains algorithms to train deep neural networks, using scans of freehand ultrasound image frames acquired with ground-truth frame locations from external spatial trackers. The aim is to reconstruct the spatial frame locations or relative transformation between them, on the newly acquired scans.

The data can be downloaded here. We have collected a new large freehand ultrasound dataset and are organising MICCAI2024&2025 Challenges TUS-REC Challenge. Check Part 1 and Part 2 of the training dataset for TUS-REC2024, and Train Data for TUS-REC2025.

Steps to run the code

1. Clone the repository.

git clone https://github.com/ucl-candi/freehand.git

2. Navigate to the root directory.

cd freehand

3. Install conda environment

conda create -n FUS python=3.9.13
conda activate FUS
pip install -r requirements.txt

4. Download data and put Freehand_US_data.zip into ./data directory. (You may need to install zenodo_get)

pip3 install zenodo_get
zenodo_get 7740734
mv Freehand_US_data.zip ./data

5. Unzip.

Unzip Freehand_US_data.zip into ./data/Freehand_US_data directory.

unzip data/Freehand_US_data.zip -d ./data

6. Make sure the data folder structure is the same as follows.

├── data/ 
│ ├── Freehand_US_data/ 
│  ├── 000/
│    ├── *.mha
│    ├── ...
│  ├── ...
│  ├── 018/ 

7. Data processing (Generate one .h5 file, using downloaded .mha files)

python data/prep.py

8. Train model

python scripts/train.py

9. Test model

python scripts/test.py

If you find this code or data set useful for your research, please consider citing some of the following works:

• Qi Li, Ziyi Shen, Qian Li, Dean C. Barratt, Thomas Dowrick, Matthew J. Clarkson, Tom Vercauteren, and Yipeng Hu. "Trackerless freehand ultrasound with sequence modelling and auxiliary transformation over past and future frames." In 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI), pp. 1-5. IEEE, 2023. doi: 10.1109/ISBI53787.2023.10230773.
• Qi Li, Ziyi Shen, Qianye Yang, Dean C. Barratt, Matthew J. Clarkson, Tom Vercauteren, and Yipeng Hu. "Nonrigid Reconstruction of Freehand Ultrasound without a Tracker." In International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 689-699. Cham: Springer Nature Switzerland, 2024. doi: 10.1007/978-3-031-72083-3_64.
• Qi Li, Ziyi Shen, Qian Li, Dean C. Barratt, Thomas Dowrick, Matthew J. Clarkson, Tom Vercauteren, and Yipeng Hu. "Long-term Dependency for 3D Reconstruction of Freehand Ultrasound Without External Tracker." IEEE Transactions on Biomedical Engineering, vol. 71, no. 3, pp. 1033-1042, 2024. doi: 10.1109/TBME.2023.3325551.
• Qi Li, Ziyi Shen, Qian Li, Dean C. Barratt, Thomas Dowrick, Matthew J. Clarkson, Tom Vercauteren, and Yipeng Hu. "Privileged Anatomical and Protocol Discrimination in Trackerless 3D Ultrasound Reconstruction." In International Workshop on Advances in Simplifying Medical Ultrasound, pp. 142-151. Cham: Springer Nature Switzerland, 2023. doi: https://doi.org/10.1007/978-3-031-44521-7_14.