Segmentation Of Brain MR Images Research Articles

Segmentation of Brain MR Images by Using Fully Convolutional Network and Gaussian Mixture Model with Spatial Constraints

Accurate segmentation of brain tissue from magnetic resonance images (MRIs) is a critical task for diagnosis, treatment, and clinical research. In this paper, a novel algorithm (GMMD-U) that incorporates the modified full convolutional neural network U-net and Gaussian-Dirichlet mixture model (GMMD) with spatial constraints is presented. The proposed GMMD-U considers the local spatial relationships by assuming that the prior probability obeys the Dirichlet distribution. Specifically, GMMD is applied for extracting brain tissue that has a distinct intensity region and modified U-net is exploited to correct the wrong-classification areas caused by GMMD or other conventional approaches. The proposed GMMD-U is designed to take advantage of the statistical model-based segmentation techniques and deep neural network. We evaluate the performance of GMMD-U on a publicly available brain MRI dataset by comparing it with several existing algorithms, and the results reported reveal that the proposed framework can accurately detect the brain tissue from MRIs. The proposed learning-based integrated framework could be effective for brain tissue segmentation, which will be helpful for surgeons in brain disease diagnosis.

Brightness preserving optimized weighted bi‐histogram equalization algorithm and its application to MR brain image segmentation

AbstractMedical image segmentation is crucial for neuroscience research and computer‐aided diagnosis. However, intensity inhomogeneity and existence of noise in magnetic resonance images lead to incorrect segmentation. In this article, an effective method called enhanced fuzzy level set algorithm is presented to segment the white matter, gray matter, and cerebrospinal fluid automatically in contrast‐enhanced brain images. In this method, first, exposure threshold is computed to divide the input histogram into two sub‐histograms of different gray levels. The input histogram is clipped using a mean gray level to control the excessive enhancement rate. Then, these two sub‐histograms are modified and equalized independently to get a better contrast enhanced image. Finally, an enhanced fuzzy level set algorithm is employed to facilitate image segmentation. The extensive experimental results proved the outstanding performance of the proposed algorithm compared with other existing methods. The results conform its effectiveness for MR brain image segmentation.

3D unsupervised modified spatial fuzzy c-means method for segmentation of 3D brain MR image

This paper proposed a novel 3D unsupervised spatial fuzzy-based brain MRI volume segmentation technique in the presence of intensity inhomogeneity and noise. Instead of static masking, dynamic 3D masking has been proposed to measure the correlation among neighbors. The local membership function is defined based on the weighted correlation among neighbors. The local and global membership functions are combined to suppress the inhomogeneity and noise at the time of clustering. A weighted function is defined based on the 3D dynamic neighborhood to optimize the objective function in 3D space. In 2D slice-based MRI image segmentation techniques, the selection of the slice of interest is very important and it depends on the experience and skills of the expertise. As the proposed unsupervised method segments the 3D brain MRI volume as a whole, there is no need of such expertise. The detailed analysis of the results shows that the proposed unsupervised spatial fuzzy-based 3D brain MRI image segmentation technique is superior over the considered state-of-the-art methods in terms of cluster validation functions.

WITHDRAWN: A Novel Robust Local Anisotropic Clustering Model for Tissue Segmentation and Bias Field Correction of Brain MR Image

This paper had been withdrawn by the authors, but has been inadvertently published by iJOE. It is hereby officially deleted from the issue.The International Journal of Online and Biomedical Engineering apologizes to the authors for the trouble caused.

A Novel Distributed Multitask Fuzzy Clustering Algorithm for Automatic MR Brain Image Segmentation.

Artificial intelligence algorithms have been used in a wide range of applications in clinical aided diagnosis, such as automatic MR image segmentation and seizure EEG signal analyses. In recent years, many machine learning-based automatic MR brain image segmentation methods have been proposed as auxiliary methods of medical image analysis in clinical treatment. Nevertheless, many problems regarding precise medical images, which cannot be effectively utilized to improve partition performance, remain to be solved. Due to the poor contrast in grayscale images, the ambiguity and complexity of MR images, and individual variability, the performance of classic algorithms in medical image segmentation still needs improvement. In this paper, we introduce a distributed multitask fuzzy c-means (MT-FCM) clustering algorithm for MR brain image segmentation that can extract knowledge common among different clustering tasks. The proposed distributed MT-FCM algorithm can effectively exploit information common among different but related MR brain image segmentation tasks and can avoid the negative effects caused by noisy data that exist in some MR images. Experimental results on clinical MR brain images demonstrate that the distributed MT-FCM method demonstrates more desirable performance than the classic signal task method.

Multiseg pipeline: automatic tissue segmentation of brain MR images with subject-specific atlases.

Automated segmentation and labeling of individual brain anatomical regions is challenging due to individual structural variability. Although, atlas-based segmentation has shown its potential for both tissue and structure segmentation, the inherent natural variability as well as disease-related changes in MR appearance is often inappropriately represented by a single atlas image. In order to have a more accurate representation, several atlases may be used for the segmentation task in a given neuroimaging study. In this paper, we present the MultisegPipeline, it uses multiple atlases that have been visually inspected and capture the expected variability in a neonatal population. The MultisegPipeline transfers the labeled regions from each atlas to the target image using deformable registration (ANTs1 or QuickSilver2 is available for this task). Additionally, the set of labels are merged using a label fusion technique that reduces the errors produced by the registration. The final output is a single label map that combines the results produced by all atlases into a consensus solution. In our study, the MultisegPipeline is used to segment brain MR images from 31 infants, a leave-one-out strategy was used to test our framework. The average dice score coefficient was 0.89.

Superpixel Segmentation of Brain MR Images Based on Probabilistic Weighted Geodesic Distance

Superpixel Segmentation of Brain MR Images Based on Probabilistic Weighted Geodesic Distance

MultisegPipeline: Automatic tissue segmentation of brain MR images with subject-specific atlases

MultisegPipeline: Automatic tissue segmentation of brain MR images with subject-specific atlases

Multi-Atlas Based Methods in Brain MR Image Segmentation.

Brain region-of-interesting (ROI) segmentation is an important prerequisite step for many computer-aid brain disease analyses. However, the human brain has the complicated anatomical structure. Meanwhile, the brain MR images often suffer from the low intensity contrast around the boundary of ROIs, large inter-subject variance and large inner-subject variance. To address these issues, many multi-atlas based segmentation methods are proposed for brain ROI segmentation in the last decade. In this paper, multi-atlas based methods for brain MR image segmentation were reviewed regarding several registration toolboxes which are widely used in the multi-atlas methods, conventional methods for label fusion, datasets that have been used for evaluating the multi-atlas methods, as well as the applications of multi-atlas based segmentation in clinical researches. We propose that incorporating the anatomical prior into the end-to-end deep learning architectures for brain ROI segmentation is an important direction in the future.

Multi-channeled MR brain image segmentation: A novel double optimization approach combined with clustering technique for tumor identification and tissue segmentation

Growth of cancer cells within the human body is a major outcome of the manipulation of cells and it has resulted in the deterioration of the life span of humans. The impact of cancer cells is irretrievable and it has paved the way to the formation of tumors within the human body. For achieving and developing a single-structured framework to prominently identify the tumor regions and segmenting the tissue structures specifically in human brain, a novel combinational algorithm is proposed through this paper. The algorithm has been embodied with two optimization techniques namely particle swarm optimization (PSO) and bacteria foraging optimization (BFO), wherein, PSO helps in finding the best position of global bacterium for BFO, consecutively, BFO supports the modified fuzzy c means (MFCM) algorithm by providing optimized cluster heads. Finally, MFCM segments the tissue regions and identifies the tumor portion, thereby reducing the interaction and complication experienced by a radiologist during patient diagnosis. The strength of the proposed algorithm is proven by comparing it with the state-of-the-art techniques by means of evaluation parameters like mean squared error (MSE), peak signal to noise ratio (PSNR), sensitivity, specificity, etc., Data sets used in this paper were exclusively obtained from hospital, Brain web simulator and BRATS-2013 challenge. The sensitivity and specificity values for 115 MR brain slice images are 0.9545 and 0.9905, which prove the segmentation ability and multitude characteristics possessed by the proposed PSBFO based MFCM (PSBFO-MFCM) algorithm.

An Accurate Automated Local Similarity Factor-Based Neural Tree Approach toward Tissue Segmentation of Newborn Brain MRI.

Segmentation of brain MR images of neonates is a primary step for assessment of brain evolvement. Advanced segmentation techniques used for adult brain MRI are not companionable for neonates, due to extensive dissimilarities in tissue properties and head structure. Existing segmentation methods for neonates utilizes brain atlases or requires manual elucidation, which results into improper and atlas dependent segmentation. The primary objective of this work is to develop fully automatic, atlas free, and robust system to segment and classify brain tissues of newborn infants from magnetic resonance images. In this study, we propose a fully automatic, atlas-free pipeline based Neural Tree approach for segmentation of newborn brain MRI which utilizes resourceful local resemblance factor such as concerning, connectivity, structure, and relative tissue location. Physical collaboration and uses of an atlas are not required in proposed method and at the same time skirting atlas-associated bias which results in improved segmentation. Proposed technique segments and classify brain tissues both at global and tissue level. We examined our results through visual assessment by neonatologists and quantitative comparisons that show first-rate concurrence with proficient manual segmentations. The implementation results of the proposed technique provided a good overall accuracy of 91.82% for the segmentation of brain tissues as compared with other methods. The pipelined-based neural tree approach along with local similarity factor segments and classify brain tissues. The proposed automated system have higher dice similarity coefficient as well as computational speed.

Survey on brain tumor segmentation and feature extraction of MR images

Brain tumor analysis plays an important role in medical imaging applications and in delivering a huge amount of anatomical and functional information, which increases and simplifies the diagnosis and disease therapy planning. However, the presence of image artifacts such as noise, intensity inhomogeneity and partial volume effect in magnetic resonance images can aggressively affect the quantitative brain tumor analysis. Also, the complex anatomy of the brain is another necessary factor to deal with. To avoid or reduce manual segmentation error, the automatic segmentation and detection of tumor have become the most challenging task for radiologists and clinicians. In this paper, most commonly used MR brain image segmentation algorithms and most popular brain MRI features are surveyed and summarized with an emphasis on their characteristics, merits, and demerits of these techniques. This paper presents a categorization of various segmentation algorithms ranging from simple threshold methods to high-level segmentation techniques such as deformable methods, graph-based, and deep learning approaches with a focus on gliomas which is most common of all malignant brain and central nervous system tumors. We also discuss the current trends with a focus on brain tumor segmentation, tissue segmentation and lesion detection using deep learning methods such as deep neural networks and convolutional neural networks. We also mentioned the future improvements to standardize the MRI-based brain tumor detection method for clinical use.

A novel approach for segmentation and classification of brain MR images using cluster deformable based fusion approach.

Segmentation of tumor form brain MR images is the most important and tedious task in the medical field. In this paper, A Cluster deformable based fusion approach which uses both deformable and K-Means clustering scheme for Segmentation is discussed. The features of tumor and non tumor cases are extracted with the use of the Power Local Binary Pattern (LBP) Operator after completion of the segmentation process. The extracted features are fed to Naive Bayes classifier to perform the process of classification. Here, the validation of the proposed system is done using standard validation methods such as accuracy, specificity, sensitivity and RoC metrics. The developed method is applied for MR images collected from standard SimBRATS database. Experimentation results shows that the proposed method performs better when compared to the traditional clustering and deformable methods and this scheme got accuracy of 84.8%.

Atlas pre-selection strategies to enhance the efficiency and accuracy of multi-atlas brain segmentation tools.

Multi-atlas brain segmentation of human brain MR images allows quantification research in structural neuroimaging. To achieve high accuracy and computational efficiency of segmentation relies on a custom subset of atlases for each target subject. However, the criterion for atlas pre-selection remains an open question. In this study, two atlas pre-selection approaches based on location-based feature matching were proposed and compared to random and mutual information-based methods using a database of 47 atlases. A varying number of atlases ranked top with hierarchical structural granularity were compared using Dice overlap. The results indicated that the proposed 4L approach consistently led to the highest level of accuracy at a given number of employed atlases in both adult and geriatric populations. In addition, the proposed two methods (4L and LV) can reduce 20 times computational time compared with the stereotypical mutual information-based method. Our pre-selection strategy would provide better segmentation performance in terms of both accuracy and efficiency. The proposed atlas pre-selection will be further implemented into our online automatic brain image segmentation system (www.mricloud.org).

Enhanced and Explored Intuitionistic Rough Based Fuzzy C-means Approach for MR Brain Image Segmentation

Segmentation of magnetic resonance images is medically complex and important for study and diagnosis of medical brain images, because of its sensitivity in terms of noise for brain medical images. These are the main issues in classification of brain images. Because of uncertainty & vagueness of brain medical images, so that rough sets, fuzzy sets and Rough sets are mathematical tools evaluate and handle uncertainty and vagueness in medical brain images. Traditionally, different type of fuzzy sets, Rough sets and rough sets based approaches were introduced, they have different several drawbacks with respect to different parameters. So this paper introduces a novel image segmentation calculation method i.e. Enhanced and Explored Intuitionistic Rough based Fuzzy C-means Approach (EEISFCMA) with estimation of weight bias parameter for brain image segmentation. Intuitionistic Rough based fuzzy sets are generalized form of fuzzy, Rough sets and their representative elements are evaluated with non-membership and membership value. Proposed algorithm of this paper consists standard features of existing clustering without spatial weight context data, it defines sensitive of noise in brain images, so that our proposed algorithm deals with intensity and noise reduction of brain image effectively. Furthermore, to reduce iterations in clustering, proposed algorithm initializes cluster centroid based on weight measure using max-dist evaluation method before execution of proposed algorithm. Experimental results of proposed approach carried out efficient image segmentation results compared to existing segmented approaches developed in brain image and other related images. Mainly proposed approach have consists better experimental evaluation based on results.

3D Randomized Connection Network with Graph-based Label Inference.

In this paper, a novel 3D deep learning network is proposed for brain MR image segmentation with randomized connection, which can decrease the dependency between layers and increase the network capacity. The convolutional LSTM and 3D convolution are employed as network units to capture the long-term and short-term 3D properties respectively. To assemble these two kinds of spatial-temporal information and refine the deep learning outcomes, we further introduce an efficient graph-based node selection and label inference method. Experiments have been carried out on two publicly available databases and results demonstrate that the proposed method can obtain competitive performances as compared with other state-of-the-art methods.

Local contextual information and Gaussian function induced fuzzy clustering algorithm for brain MR image segmentation and intensity inhomogeneity estimation

This paper presents a fuzzy clustering algorithm, where local contextual information and a Gaussian function are incorporated into the objective function, for simultaneous brain MR image segmentation and intensity inhomogeneity estimation. In doing so, for each pixel, we define a local contextual information, which actually defines its association among the other neighboring pixels based on intensity distribution. In particular, this information defines the possibility of the pixel to belong into a specific tissue type. Whereas, for each tissue region, a Gaussian surface is defined to estimate the intensity inhomogeneity (IIH) using the local image gradients, which are believed to be caused by the IIH. We use this Gaussian surface to compensate the effect of IIH. In addition, for each pixel, we have introduced global and local membership functions, which in combined in association with the other parameters are responsible for generation of cluster prototypes. The IIH of the entire image region is iteratively removed from the image and the final segmentation result is obtained based on the global membership values. The simulation results on two benchmarks brain MR image databases and four volumes of real-patient brain MR image data show its efficiency and superiority over other fuzzy-based clustering algorithms.

Supervoxel based method for multi-atlas segmentation of brain MR images

Multi-atlas segmentation has been widely applied to the analysis of brain MR images. However, the state-of-the-art techniques in multi-atlas segmentation, including both patch-based and learning-based methods, are strongly dependent on the pairwise registration or exhibit huge spatial inconsistency. The paper proposes a new segmentation framework based on supervoxels to solve the existing challenges of previous methods. The supervoxel is an aggregation of voxels with similar attributes, which can be used to replace the voxel grid. By formulating the segmentation as a tissue labeling problem associated with a maximum-a-posteriori inference in Markov random field, the problem is solved via a graphical model with supervoxels being considered as the nodes. In addition, a dense labeling scheme is developed to refine the supervoxel labeling results, and the spatial consistency is incorporated in the proposed method. The proposed approach is robust to the pairwise registration errors and of high computational efficiency. Extensive experimental evaluations on three publically available brain MR datasets demonstrate the effectiveness and superior performance of the proposed approach.

MR Image Segmentation Based on Contrast Enhancement with Collaborative Learning

Magnetic resonance imaging (MRI) has been widely used on clinical applications. Due to different weighting strategies MRI, different weighting images have different contrast even for the same anatomy structure of the same subject. Since some MR brain images have low contrast, different types of tissues such as white matter, grey matter, and cerebrospinal fluid (CSF) are difficult to be divided and segmented. Image contrast needs to be enhanced for better post-processing and image analysis. In this paper, a two-step MR brain image segmentation technique is proposed to solve low contrast MR image segmentation problem. A collaborative learning based image enhancement is firstly applied on low contrast MR brain image. Then, spectral clustering algorithm is used for segmenting enhanced image. Experimental results illustrate that the proposed 2-step segmentation method is able to identify boundaries between tissues well, so that MR image segmentation accuracy is improved in compared to image segmentation without contrast enhancement and exact histogram equalization enhanced image segmentation.

Fusion of Gaussian Mixture Model and Spatial Fuzzy C-Means for Brain MR Image Segmentation

Brain image segmentation into white matter, grey matter and cerebrospinal fluid is a very popular yet challenging area in medical image processing. The Fuzzy C-Means clustering algorithm is quite used because of its ability to ensure multiple member of pixels in several clusters. This is further appreciated when the spatial information of the data is considered, as such, the algorithm becomes much more robust to noise. However, pixels which form part of the non-overlapped tissues often come out inaccurate. To solve this problem, a spatial fuzzy algorithm fused with the Gaussian mixture model using the expectation maximization algorithm is presented in this paper. The proposed algorithm is a fusion of the Fuzzy C-Means and gaussian Mixture Model algorithms for segmenting tissues having multiple cluster memembership as well as single clustering. The results of the proposed algorithm are compared with the other classical algorithms against the manually segmented results of the input image. The estimated results of the proposed algorithm using the Dice and Jaccard similarity index indicates improved accuracy to the other algorithms.