RESUMO
Brain tissue segmentation from MR images is a critical step for quantifying the brain morphology in neuroimaging studies. While deep learning (DL) based brain tissue segmentation methods have achieved promising performance, most of them are built upon supervised learning and therefore their performance is bounded by the training data used and limited by the small size of datasets with manual segmentation labels. To leverage the large amount of unlabeled brain imaging data, we develop an unsupervised DL model for joint brain tissue segmentation and bias field estimation using cascaded convolutional networks. The proposed DL model consists of multiple cascaded bias field estimation modules and one segmentation module. The bias field estimation modules are applied to the input image for estimating the bias field and generating a bias-free image recursively, and the bias field corrected image is then fed into the segmentation module to obtain the brain tissue segmentation result. A Gaussian mixture model is adopted to characterize the bias-free image with tissue-specific intensity statistics and the model fitting error is adopted as the loss function to guide the optimization of the model parameters progressively in an unsupervised setting. We have evaluated the proposed method on the HCP-Aging and HCP-Development datasets. Quantitative results have demonstrated that our unsupervised DL model could obtain competitive bias field correction and segmentation performance, compared with state-of-the-art bias field correction methods and unsupervised segmentation methods.
RESUMO
BACKGROUND: Intensity inhomogeneity is one of the common artifacts in image processing. This artifact makes image segmentation more challenging and adversely affects the performance of intensity-based image processing algorithms. NEW METHOD: In this paper, a novel region-based level set method is proposed for segmenting the images with intensity inhomogeneity with applications to brain tumor segmentation in magnetic resonance imaging (MRI) scans. For this purpose, the inhomogeneous regions are first modeled as Gaussian distributions with different means and variances, and then transferred into a new domain, where preserves the Gaussian intensity distribution of each region but with better separation. Moreover, our method can perform bias field correction. To this end, the bias field is represented by a linear combination of smooth base functions that enables better intensity inhomogeneity modeling. Therefore, level set fundamental formulation and bias field are modified in the proposed approach. RESULTS: To assess the performance of the proposed method, different inhomogeneous images, including synthetic images as well as real brain magnetic resonance images from BraTS 2017 dataset are segmented. Being evaluated by Dice, Jaccard, Sensitivity, and Specificity metrics, the results show that the proposed method suppresses the side effect of the over-smoothing object boundary and it has good accuracy in the segmentation of images with extreme intensity non-uniformity. The mean values of these metrics in brain tumor segmentation are 0.86 ± 0.03, 0.77 ± 0.05, 0.94 ± 0.04, 0.99 ± 0.003, respectively. COMPARISON WITH EXISTING METHOD(S): Our method were compared with six state-of-the-art image segmentation methods: Chan-Vese (CV), Local Intensity Clustering (LIC), Local iNtensity Clustering (LINC), Global inhomogeneous intensity clustering (GINC), Multiplicative Intrinsic Component Optimization (MICO), and Local Statistical Active Contour Model (LSACM) models. We used qualitative and quantitative comparison methods for segmenting synthetic and real images. Experiments indicate that our proposed method is robust to noise and intensity non-uniformity and outperforms other state-of-the-art segmentation methods in terms of bias field correction, noise resistance, and segmentation accuracy. CONCLUSIONS: Experimental results show that the proposed model is capable of accurate segmentation and bias field estimation simultaneously. The proposed model suppresses the side effect of the over-smoothing object boundary. Moreover, our model has good accuracy in the segmentation of images with extreme intensity non-uniformity.
Assuntos
Neoplasias Encefálicas , Imageamento por Ressonância Magnética , Algoritmos , Encéfalo/diagnóstico por imagem , Neoplasias Encefálicas/diagnóstico por imagem , Humanos , Processamento de Imagem Assistida por Computador , NeuroimagemRESUMO
In this paper, we extend the multiplicative intrinsic component optimization (MICO) algorithm to multichannel MR image segmentation, with focus on segmentation of multiple sclerosis (MS) lesions. The MICO algorithm was originally proposed by Li et al. in Ref. [1] for normal brain tissue segmentation and intensity inhomogeneity correction of a single channel MR image, which exhibits desirable advantages over other methods for MR image segmentation and intensity inhomogeneity correction in terms of segmentation accuracy and robustness. In this paper, we extend the MICO algorithm to multi-channel MR image segmentation and enable the segmentation of MS lesions. We assign different weights for different channels to control the impact of each channel. The weighted channels allow the enhancement of the impact of the FLAIR image on the segmentation of MS lesions by assigning a larger weight to the FLAIR image channel than the other channels. With the inherent mechanism of estimation of the bias field, our method is able to deal with the intensity inhomogeneity in the input multi-channel MR images. In the application of our method, we only use T1-w and FLAIR images as the input two channel MR images. Experimental results show promising result of our method.
Assuntos
Encéfalo/diagnóstico por imagem , Imageamento por Ressonância Magnética/métodos , Esclerose Múltipla/diagnóstico por imagem , Algoritmos , Encéfalo/patologia , Humanos , Processamento de Imagem Assistida por Computador , Esclerose Múltipla/patologiaRESUMO
This paper proposes a new energy minimization method called multiplicative intrinsic component optimization (MICO) for joint bias field estimation and segmentation of magnetic resonance (MR) images. The proposed method takes full advantage of the decomposition of MR images into two multiplicative components, namely, the true image that characterizes a physical property of the tissues and the bias field that accounts for the intensity inhomogeneity, and their respective spatial properties. Bias field estimation and tissue segmentation are simultaneously achieved by an energy minimization process aimed to optimize the estimates of the two multiplicative components of an MR image. The bias field is iteratively optimized by using efficient matrix computations, which are verified to be numerically stable by matrix analysis. More importantly, the energy in our formulation is convex in each of its variables, which leads to the robustness of the proposed energy minimization algorithm. The MICO formulation can be naturally extended to 3D/4D tissue segmentation with spatial/sptatiotemporal regularization. Quantitative evaluations and comparisons with some popular softwares have demonstrated superior performance of MICO in terms of robustness and accuracy.
Assuntos
Algoritmos , Artefatos , Encéfalo/anatomia & histologia , Aumento da Imagem/métodos , Interpretação de Imagem Assistida por Computador/métodos , Imageamento por Ressonância Magnética/métodos , Reconhecimento Automatizado de Padrão/métodos , Humanos , Reprodutibilidade dos Testes , Sensibilidade e EspecificidadeRESUMO
Segmentation of multiple sclerosis (MS) lesion is important for many neuroimaging studies. In this paper, we propose a novel algorithm for automatic segmentation of MS lesions from multi-channel MR images (T1W, T2W and FLAIR images). The proposed method is an extension of Li et al.'s algorithm in [1], which only segments the normal tissues from T1W images. The proposed method is aimed to segment MS lesions, while normal tissues are also segmented and bias field is estimated to handle intensity inhomogeneities in the images. Another contribution of this paper is the introduction of a nonlocal means technique to achieve spatially regularized segmentation, which overcomes the influence of noise. Experimental results have demonstrated the effectiveness and advantages of the proposed algorithm.