Zhang Jinwei, Liu Zhe, Zhang Shun, Zhang Hang, Spincemaille Pascal, Nguyen Thanh D, Sabuncu Mert R, Wang Yi
Department of Biomedical Engineering, Cornell University, Ithaca, NY, USA; Department of Radiology, Weill Medical College of Cornell University, New York, NY, USA.
Department of Radiology, Weill Medical College of Cornell University, New York, NY, USA.
Neuroimage. 2020 May 1;211:116579. doi: 10.1016/j.neuroimage.2020.116579. Epub 2020 Jan 22.
Deep learning (DL) is increasingly used to solve ill-posed inverse problems in medical imaging, such as reconstruction from noisy and/or incomplete data, as DL offers advantages over conventional methods that rely on explicit image features and hand engineered priors. However, supervised DL-based methods may achieve poor performance when the test data deviates from the training data, for example, when it has pathologies not encountered in the training data. Furthermore, DL-based image reconstructions do not always incorporate the underlying forward physical model, which may improve performance. Therefore, in this work we introduce a novel approach, called fidelity imposed network edit (FINE), which modifies the weights of a pre-trained reconstruction network for each case in the testing dataset. This is achieved by minimizing an unsupervised fidelity loss function that is based on the forward physical model. FINE is applied to two important inverse problems in neuroimaging: quantitative susceptibility mapping (QSM) and under-sampled image reconstruction in MRI. Our experiments demonstrate that FINE can improve reconstruction accuracy.
深度学习（DL）越来越多地用于解决医学成像中的不适定逆问题，例如从噪声和/或不完整数据进行重建，因为与依赖显式图像特征和手工设计先验的传统方法相比，深度学习具有优势。然而，当测试数据与训练数据存在偏差时，例如测试数据具有训练数据中未遇到的病变时，基于监督深度学习的方法可能表现不佳。此外，基于深度学习的图像重建并不总是纳入潜在的正向物理模型，而纳入该模型可能会提高性能。因此，在这项工作中，我们引入了一种新颖的方法，称为保真度强制网络编辑（FINE），它针对测试数据集中的每个案例修改预训练重建网络的权重。这是通过最小化基于正向物理模型的无监督保真度损失函数来实现的。FINE被应用于神经成像中的两个重要逆问题：定量磁化率映射（QSM）和MRI中的欠采样图像重建。我们的实验表明，FINE可以提高重建精度。
