Magnetic resonance imaging (MRI) is a principal radiological modality that provides radiation-free, abundant, and diverse information about the whole human body for medical diagnosis, but suffers from prolonged scan time. The scan time can be significantly reduced through k-space undersampling but the introduced artifacts need to be removed in image reconstruction. Although deep learning (DL) has emerged as a powerful tool for image reconstruction in fast MRI, its potential in multiple imaging scenarios remains largely untapped. This is because not only collecting large-scale and diverse realistic training data is generally costly and privacy-restricted, but also existing DL methods are hard to handle the practically inevitable mismatch between training and target data. Here, we present a Physics-Informed Synthetic data learning framework for Fast MRI, called PISF, which is the first to enable generalizable DL for multi-scenario MRI reconstruction using solely one trained model. For a 2D image, the reconstruction is separated into many 1D basic problems and starts with the 1D data synthesis, to facilitate generalization. We demonstrate that training DL models on synthetic data, integrated with enhanced learning techniques, can achieve comparable or even better in vivo MRI reconstruction compared to models trained on a matched realistic dataset, reducing the demand for real-world MRI data by up to 96%. Moreover, our PISF shows impressive generalizability in multi-vendor multi-center imaging. Its excellent adaptability to patients has been verified through 10 experienced doctors' evaluations. PISF provides a feasible and cost-effective way to markedly boost the widespread usage of DL in various fast MRI applications, while freeing from the intractable ethical and practical considerations of in vivo human data acquisitions.
Magnetic Resonance Spectroscopy (MRS) is an important non-invasive technique for in vivo biomedical detection. However, it is still challenging to accurately quantify metabolites with proton MRS due to three problems: Serious overlaps of metabolite signals, signal distortions due to non-ideal acquisition conditions and interference with strong background signals including macromolecule signals. The most popular software, LCModel, adopts the non-linear least square to quantify metabolites and addresses these problems by introducing regularization terms, imperfection factors of non-ideal acquisition conditions, and designing several empirical priors such as basissets of both metabolites and macromolecules. However, solving such a large non-linear quantitative problem is complicated. Moreover, when the signal-to-noise ratio of an input MRS signal is low, the solution may have a large deviation. In this work, deep learning is introduced to reduce the complexity of solving this overall quantitative problem. Deep learning is designed to predict directly the imperfection factors and the overall signal from macromolecules. Then, the remaining part of the quantification problem becomes a much simpler effective fitting and is easily solved by Linear Least Squares (LLS), which greatly improves the generalization to unseen concentration of metabolites in the training data. Experimental results show that compared with LCModel, the proposed method has smaller quantification errors for 700 sets of simulated test data, and presents more stable quantification results for 20 sets of healthy in vivo data at a wide range of signal-to-noise ratio. Qnet also outperforms other deep learning methods in terms of lower quantification error on most metabolites. Finally, QNet has been deployed on a cloud computing platform, CloudBrain-MRS, which is open accessed at https://csrc.xmu.edu.cn/CloudBrain.html.
In this paper, we surveyed the existing literature studying different approaches and algorithms for the four critical components in the general branch and bound (B&B) algorithm, namely, branching variable selection, node selection, node pruning, and cutting-plane selection. However, the complexity of the B&B algorithm always grows exponentially with respect to the increase of the decision variable dimensions. In order to improve the speed of B&B algorithms, learning techniques have been introduced in this algorithm recently. We further surveyed how machine learning can be used to improve the four critical components in B&B algorithms. In general, a supervised learning method helps to generate a policy that mimics an expert but significantly improves the speed. An unsupervised learning method helps choose different methods based on the features. In addition, models trained with reinforcement learning can beat the expert policy, given enough training and a supervised initialization. Detailed comparisons between different algorithms have been summarized in our survey. Finally, we discussed some future research directions to accelerate and improve the algorithms further in the literature.