A Deep Network for Joint Registration and Parcellation of Cortical Surfaces

A weakly supervised NMF method to decipher molecular subtype-related dynamic patterns in breast DCE-MR images

Objective. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is an important imaging modality for breast cancer diagnosis. Intratumoral heterogeneity causes a major challenge in the interpretation of breast DCE-MRI. Previous studies have introduced decomposition methods on DCE-MRI to reveal intratumoral heterogeneity by analyzing distinct dynamic patterns within each tumor. However, these methods estimated the dynamic patterns and their corresponding component coefficients in an unsupervised manner, without considering any clinically relevant information.Approach. To decipher molecular subtype-related dynamic patterns, we propose a weakly supervised nonnegative matrix factorization method (WSNMF), which is able to decompose the pixel kinetics of DCE-MRI with image-level subtype labels. The WSNMF is developed based on a discriminant nonnegative matrix factorization (NMF) to utilize coarse-grained subtype information, in which between- and within-class scatters are defined on the mean vector of component coefficients over all pixels in each tumor, rather than directly on the vector of component coefficients of each pixel.Main results. Experiments demonstrated that the dynamic patterns identified by WSNMF had superior performance in distinguishing between luminal A and the other subtype tumors. The classification performance was evaluated using the area under the receiver operating characteristic curve (AUC). WSNMF yielded better classification performance (AUC = 0.822) than other heterogeneity analysis methods, including two partitioning-based methods (KPC with AUC = 0.697 and TTP with AUC = 0.760) and two unsupervised decomposition-based methods (PCA with AUC = 0.774 and NMF with AUC = 0.797).Significance. Our method adds a valuable new perspective into DCE-MRI decomposition-based heterogeneity analysis by taking advantage of intrinsic tumor characteristics to improve the diagnosis of breast cancer.

A review of deep learning-based deformable medical image registration

The alignment of images through deformable image registration is vital to clinical applications (e.g., atlas creation, image fusion, and tumor targeting in image-guided navigation systems) and is still a challenging problem. Recent progress in the field of deep learning has significantly advanced the performance of medical image registration. In this review, we present a comprehensive survey on deep learning-based deformable medical image registration methods. These methods are classified into five categories: Deep Iterative Methods, Supervised Methods, Unsupervised Methods, Weakly Supervised Methods, and Latest Methods. A detailed review of each category is provided with discussions about contributions, tasks, and inadequacies. We also provide statistical analysis for the selected papers from the point of view of image modality, the region of interest (ROI), evaluation metrics, and method categories. In addition, we summarize 33 publicly available datasets that are used for benchmarking the registration algorithms. Finally, the remaining challenges, future directions, and potential trends are discussed in our review.

Fast Spherical Mapping of Cortical Surface Meshes Using Deep Unsupervised Learning

Spherical mapping of cortical surface meshes provides a more convenient and accurate space for cortical surface registration and analysis and thus has been widely adopted in neuroimaging field. Conventional approaches typically first inflate and project the original cortical surface mesh onto a sphere to generate an initial spherical mesh which contains large distortions. Then they iteratively reshape the spherical mesh to minimize the metric (distance), area or angle distortions. However, these methods suffer from two major issues: 1) the iterative optimization process is computationally expensive, making them not suitable for large-scale data processing; 2) when metric distortion cannot be further minimized, either area or angle distortion is minimized at the expense of the other, which is not flexible to generate application-specific meshes based on both of them. To address these issues, for the first time, we propose a deep learning-based algorithm to learn the mapping between the original cortical surface and spherical surface meshes. Specifically, we take advantage of the Spherical U-Net model to learn the spherical diffeomorphic deformation field for minimizing the distortions between the icosahedron-reparameterized original surface and spherical surface meshes. The end-to-end unsupervised learning scheme is very flexible to incorporate various optimization objectives. We further integrate it into a coarse-to-fine multi-resolution framework for better correcting fine-scaled distortions. We have validated our method on 800+ cortical surfaces, demonstrating reduced distortions than FreeSurfer (the most popularly used tool), while speeding up the process from 20 min to 5 s.