An Efficient Computational Approach for the Detection of MR Brain Tissues in the Presence of Noise and Intensity Inhomogeneity

Author(s):  
G. Sandhya ◽  
Kande Giri Babu ◽  
T. Satya Savithri

The automatic detection of brain tissues such as White Matter (WM), Gray Matter (GM), and Cerebrospinal Fluid (CSF) from the MR images of the brain using segmentation is of immense interest for the early detection and diagnosing various brain-related diseases. MR imaging technology is one of the best and most reliable ways of studying the brain. Segmentation of MR images is a challenging task due to various artifacts such as noise, intensity inhomogeneity, partial volume effects and elemental texture of the image. This work proposes a region based, efficient and modern energy minimization process called as Anisotropic Multiplicative Intrinsic Component Optimization (AMICO) for the brain image segmentation in the presence of noise and intensity inhomogeneity to separate different tissues. This algorithm uses an efficient Anisotropic diffusion filter to decrease the noise. The denoised image gets segmented after the correction of intensity inhomogeneity by the MICO algorithm. The algorithm decomposes the MR brain image as two multiplicative intrinsic components, called as the component of the true image which represents the physical properties of the brain tissue and the component of bias field that is related to intensity inhomogeneity. By optimizing the values of these two components using an efficient energy minimization technique, correction of intensity inhomogeneity and segmentation of the tissues can be achieved simultaneously. Performance evaluation and the comparison with some existing methods have validated the remarkable performance of AMICO in terms of efficiency of segmentation of brain images in the presence of noise and intensity inhomogeneity.

2017 ◽  
pp. 115-130
Author(s):  
Vijay Kumar ◽  
Jitender Kumar Chhabra ◽  
Dinesh Kumar

Image segmentation plays an important role in medical imaging applications. In this chapter, an automatic MRI brain image segmentation framework using gravitational search based clustering technique has been proposed. This framework consists of two stage segmentation procedure. First, non-brain tissues are removed from the brain tissues using modified skull-stripping algorithm. Thereafter, the automatic gravitational search based clustering technique is used to extract the brain tissues from the skull stripped image. The proposed algorithm has been applied on four simulated T1-weighted MRI brain images. Experimental results reveal that proposed algorithm outperforms the existing techniques in terms of the structure similarity measure.


Author(s):  
Vijay Kumar ◽  
Jitender Kumar Chhabra ◽  
Dinesh Kumar

Image segmentation plays an important role in medical imaging applications. In this chapter, an automatic MRI brain image segmentation framework using gravitational search based clustering technique has been proposed. This framework consists of two stage segmentation procedure. First, non-brain tissues are removed from the brain tissues using modified skull-stripping algorithm. Thereafter, the automatic gravitational search based clustering technique is used to extract the brain tissues from the skull stripped image. The proposed algorithm has been applied on four simulated T1-weighted MRI brain images. Experimental results reveal that proposed algorithm outperforms the existing techniques in terms of the structure similarity measure.


The segmentation procedure might cause error in diagnosing MR images due to the artifacts and noises that exist in it. This may lead to misclassifying normal tissue as abnormal tissue. In addition, it is also required to model the ontogenesis of a tumour, as they propagate at distinctive rates in contrast to their surroundings. Hence, it is still a challenging task to segment MR brain images due to possible noise presence, bias field and impact of partial volume. This article presents an efficient approach for segmenting MR brain images using a modified kernel based fuzzy clustering (MKFC) algorithm. In addition, this approach computes the weight of each picture element based on the local mutation coefficient (LMC). The proposed system would reflexively group normal tissues like white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF) respectively, from abnormal tissue, such as a tumour region, in MR brain images. Simulation outcomes have shown that the performance of the proposed segmentation approach is superior to the existing segmentation algorithms in terms of both ocular and quantitative analysis


2018 ◽  
Vol 7 (4.10) ◽  
pp. 197
Author(s):  
Maryjo M George ◽  
Kalaivani S

Intensity inhomogeneity is an artifact in MR brain images and causes intensity variation of same tissues on the basis of location of the tissue within the image. It is crucial to minimize this phenomenon to improve the accuracy of the computer-aided diagnosis. Unlike the several methods proposed in the past to minimize intensity inhomogeneity, this proposed method uses a pyramidal decomposition strategy to estimate the bias field in MR brain images. The bias field estimated from the proposed multi-scale framework can be effectively used for intensity inhomogeneity correction of the acquired MR data. The proposed methodology has been tested on simulated database and quantitative analyses in terms of coefficient of variation in grey matter and white matter tissue regions separately and combined coefficient of joint variation are assessed. The qualitative and quantitative analyses on the corrected data indicate that the method is effective for intensity inhomogeneity on brain MR images.  


Author(s):  
Sandhya Gudise ◽  
Giri Babu Kande ◽  
T. Satya Savithri

This paper proposes an advanced and precise technique for the segmentation of Magnetic Resonance Image (MRI) of the brain. Brain MRI segmentation is to be familiar with the anatomical structure, to recognize the deformities, and to distinguish different tissues which help in treatment planning and diagnosis. Nature’s inspired population-based evolutionary algorithms are extremely popular for a wide range of applications due to their best solutions. Teaching Learning Based Optimization (TLBO) is an advanced population-based evolutionary algorithm designed based on Teaching and Learning process of a classroom. TLBO uses common controlling parameters and it won’t require algorithm-specific parameters. TLBO is more appropriate to optimize the real variables which are fuzzy valued, computationally efficient, and does not require parameter tuning. In this work, the pixels of the brain image are automatically grouped into three distinct homogeneous tissues such as White Matter (WM), Gray Matter (GM), and Cerebro Spinal Fluid (CSF) using the TLBO algorithm. The methodology includes skull stripping and filtering in the pre-processing stage. The outcomes for 10 MR brain images acquired by utilizing the proposed strategy proved that the three brain tissues are segmented accurately. The segmentation outputs are compared with the ground truth images and high values are obtained for the measure’s sensitivity, specificity, and segmentation accuracy. Four different approaches, namely Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Bacterial Foraging Algorithm (BFA), and Electromagnetic Optimization (EMO) are likewise implemented to compare with the results of the proposed methodology. From the results, it can be proved that the proposed method performed effectively than the other.


Magnetic resonance imaging (MRI) is an incredible testing method which provides appropriate anatomical images of the body. For the diagnosis, high resolution MR images are essential to extract the detailed information about the diseases. However, with the measured MR images it’s a challenging issue in extracting the detailed information associated to disease for the posterior analysis or treatment. Usually to improve the resolution of the MR image, histogram equalization process has to be applied. In this paper, interpolation method is applied to improve the resolution of MR brain images for the histogram-ed images. And for the assessment of the skillfulness of introduced method, performance metrics such as peak signal to noise ratio (PSNR) and absolute mean brightness error (AMBE) are measured. The peak of signal for the enhanced images through interpolation will be much better and may have the good variation to the mean brightness error. With this there can be potential to the artificial intelligence for better diagnosis in complex decisive instances


In general, two risk factors such as alcohol expectancy and impulsivity have been concerned with alcohol abuse Currently, many people have been addicted to alcoholism and have an Alcohol Use Disorder (AUD) that affects neurons behavior in the human brain. Still, how such risk factors interrelate to estimate the alcoholism. To solve this problem, Fuzzy C-Regression based Alcoholism Detection (FCRAD) method has been proposed that segments the Region-Of-Interests (ROIs) from the human brain image to predict the Gray Matter Volume (GMV) reduction in the right posterior insula in women and the left thalamus in both men and women efficiently. However, it requires the detection of GMV reduction in the other brain image regions. This multi-modality can decrease the fuzziness of the partition and the crisp membership degrees were not derived easily. Therefore in this article, the GMV reduction in other regions of the brain images including right posterior insula in women and left thalamus in both men and women has been detected, an Improved FCRAD (IFCRAD) method is proposed to simplify the segmentation of the brain images by considering the second regularization term in the objective function of the FCR to take into account the noisy data. Also, the Euclidean distance is replaced with the Voronoi distance for computing different fuzzy membership functions. Moreover, new error measure and reward function are used in the objective function of the FCR to reward nearly crisp membership functions and to obtain more crisp partition. So, the brain images are segmented into gray-matter images that derive the ROIs to analyze the GMV reduction with less complexity. Finally, the experimental results illustrate the proposed IFCRAD method achieves higher accuracy than the existing AD methods.


In this research, an automated and customized neoplasm segmentation methodology is given and valid against ground truth applying simulated T1-weighted resonance pictures in twenty five subjects. a replacement intensity-based segmentation technique known as bar graph primarily based gravitational optimization algorithm is developed to segment the brain image into discriminative sections (segments) with high accuracy. whereas the mathematical foundation of this rule is given in details, the appliance of the projected rule within the segmentation of single T1-weighted pictures (T1-w) modality of healthy and lesion MR images is additionally given. The results show that the neoplasm lesion is divided from the detected lesion slice with eighty nine.6% accuracy..


2021 ◽  
Vol 7 (2) ◽  
pp. 763-766
Author(s):  
Sreelakshmi Shaji ◽  
Nagarajan Ganapathy ◽  
Ramakrishnan Swaminathan

Abstract Alzheimer’s Disease (AD) is an irreversible progressive neurodegenerative disorder. Magnetic Resonance (MR) imaging based deep learning models with visualization capabilities are essential for the precise diagnosis of AD. In this study, an attempt has been made to categorize AD and Healthy Controls (HC) using structural MR images and an Inception-Residual Network (ResNet) model. For this, T1- weighted MR brain images are acquired from a public database. These images are pre-processed and are applied to a two-layer Inception-ResNet-A model. Additionally, Gradient weighted Class Activation Mapping (Grad-CAM) is employed to visualize the significant regions in MR images identified by the model for AD classification. The network performance is validated using standard evaluation metrics. Results demonstrate that the proposed Inception-ResNet model differentiates AD from HC using MR brain images. The model achieves an average recall and precision of 69%. The Grad- CAM visualization identified lateral ventricles in the mid-axial slice as the most discriminative brain regions for AD classification. Thus, the computer aided diagnosis study could be useful in the visualization and automated analysis of AD diagnosis with minimal medical expertise.


Sign in / Sign up

Export Citation Format

Share Document