MoDL-QSM: Model-based deep learning for quantitative susceptibility mapping
Ruimin Feng,
Jiayi Zhao,
He Wang,
Baofeng Yang,
Jie Feng,
Yuting Shi,
Ming Zhang,
Chunlei Liu,
Yuyao Zhang,
Jie Zhuang,
Hongjiang Wei
Affiliations
Ruimin Feng
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Jiayi Zhao
School of Psychology, Shanghai University of Sport, Shanghai, China
He Wang
Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China
Baofeng Yang
Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China
Jie Feng
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Yuting Shi
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Ming Zhang
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Chunlei Liu
Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, CA, USA; Helen Wills Neuroscience Institute, University of California, Berkeley, CA, USA
Yuyao Zhang
School of Information and Science and Technology, ShanghaiTech University, Shanghai, China
Jie Zhuang
School of Psychology, Shanghai University of Sport, Shanghai, China
Hongjiang Wei
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China; Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, China; Corresponding author at: School of Biomedical Engineering, Shanghai Jiao Tong University, 1954 Huashan Rd, MED-X Research Institute, Shanghai, 200030, China.
Quantitative susceptibility mapping (QSM) has demonstrated great potential in quantifying tissue susceptibility in various brain diseases. However, the intrinsic ill-posed inverse problem relating the tissue phase to the underlying susceptibility distribution affects the accuracy for quantifying tissue susceptibility. Recently, deep learning has shown promising results to improve accuracy by reducing the streaking artifacts. However, there exists a mismatch between the observed phase and the theoretical forward phase estimated by the susceptibility label. In this study, we proposed a model-based deep learning architecture that followed the STI (susceptibility tensor imaging) physical model, referred to as MoDL-QSM. Specifically, MoDL-QSM accounts for the relationship between STI-derived phase contrast induced by the susceptibility tensor terms (χ13, χ23 and χ33) and the acquired single-orientation phase. The convolutional neural networks are embedded into the physical model to learn a regularization term containing prior information. χ33 and phase induced by χ13 and χ23 terms were used as the labels for network training. Quantitative evaluation metrics were compared with recently developed deep learning QSM methods. The results showed that MoDL-QSM achieved superior performance, demonstrating its potential for future applications.
