Simultaneous Segmentation and Classification of Bone Surfaces from Ultrasound Using a Multi-feature Guided CNN
The proposed MFG-CNN are implemented in Python using Pytorch. It can be trainied following details in the paper.
The dataset (including computed BSE, LPT and LP) will be released as soon as ethics approvals are obtatined.
If you use this code in your research, please consider citing our paper:
@inproceedings{wang2018simultaneous,
title={Simultaneous segmentation and classification of bone surfaces from ultrasound using a multi-feature guided CNN},
author={Wang, Puyang and Patel, Vishal M and Hacihaliloglu, Ilker},
booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
pages={134--142},
year={2018},
organization={Springer}
}