Multi-Atlas Segmentation of MR Tumor Brain Images Using Low-Rank Based Image Recovery

Abstract

We introduce a new multi-atlas segmentation (MAS) framework for MR tumor brain images. The basic idea of MAS is to register and fuse label information from multiple normal brain atlases to a new brain image for segmentation. Many MAS methods have been proposed with success. However, most of them are developed for normal brain images, and tumor brain images usually pose a great challenge for them. This is because tumors cause difficulties in registration of normal brain atlases to the tumor brain image. To address this challenge, in the first step of our MAS framework, a new low-rank method is used to get the recovered image of normal-looking brain from the MR tumor brain image based on the information of normal brain atlases. Different from conventional low-rank methods that produce the recovered image with distorted normal brain regions, our low-rank method harnesses a spatial constraint to get the recovered image with preserved normal brain regions. Then in the second step, normal brain atlases can be registered to the recovered image without influence from tumors. These two steps are iteratively proceeded until convergence, for obtaining the final segmentation of the tumor brain image. During the iteration, both the recovered image and the registration of normal brain atlases to the recovered image are gradually refined. We have compared our proposed method with a state-of-the-art method by using both synthetic and real MR tumor brain images. Experimental results show that our proposed method can get effectively recovered images and also improves segmentation accuracy.

An Indirect Multimodal Image Registration and Completion Method Guided by Image Synthesis

“…Several manually crafted metrics are frequently used, such as the sum of squared differences (SSD), cross-correlation (CC) [ 24 ], mutual information (MI) [ 25 ], normalized cross-correlation (NCC), and normalized mutual information (NMI). The optimization algorithms are mostly intensity-based [ 26 , 27 ] and feature-based [ 28 – 30 ]. Actually, image registration generally includes linear (rigid) registration and deformable (nonrigid) registration, where linear registration intends to globally align the two images, and deformable registration is used to correct local deformations.…”

Subcortical Brain Segmentation Based on Atlas Registration and Linearized Kernel Sparse Representative Classifier

“…(Z e − min) × 255 max − min (21) whereZ e is the normalized value of voxel e. max and min are maximum and minimum intensities in MRI. In the AR+LKSRC based segmentation framework for subcortical brain segmentation, we used the well-known dice similarity coefficient (DSC) [37] and recall (RC) for evaluation.…”

“…Coupé et al proposed nonlocal patch-based method for hippocampus and ventricle segmentation [17]. Typical methods with patch reconstruction are sparse representative classifier (SRC) [20], [21] and dictionary learning (DL) [22], [23]. Tong et al [23] improved discriminative dictionary learning method by designing a fixed dictionary for hippocampus segmentation.…”

Low-Rank Based Image Analyses for Pathological MR Image Segmentation and Recovery

“…However, there is lack of deep learning based methods developed for pathological medical image recovery. In contrast, the low-rank and sparse decomposition (LSD) (Wright et al, 2009; Candès et al, 2011) scheme, learning normal image appearance from unlabeled population data, has been widely employed to decompose pathological MR brain images into recovered normal brain appearances and pathological regions (Liu et al, 2015; Tang et al, 2018).…”

“…Although the low-rank and sparse analyses of computational brain tumor segmentation has attracted considerable attention during last decade, it remains several challenges. First, conventional LSD methods have to be computed on a series of aligned images (Otazo et al, 2015; Tang et al, 2018), because the image misalignment causes undesired structure differences that would interfere the representation of sparse component. Thus, the image alignment should be conducted before/during the LSD computation; however, the image alignment itself is a challenging task.…”