Brain Image Segmentation Research Articles

Differences Between MR Brain Region Segmentation Methods: Impact on Single-Subject Analysis.

For the segmentation of magnetic resonance brain images into anatomical regions, numerous fully automated methods have been proposed and compared to reference segmentations obtained manually. However, systematic differences might exist between the resulting segmentations, depending on the segmentation method and underlying brain atlas. This potentially results in sensitivity differences to disease and can further complicate the comparison of individual patients to normative data. In this study, we aim to answer two research questions: 1) to what extent are methods interchangeable, as long as the same method is being used for computing normative volume distributions and patient-specific volumes? and 2) can different methods be used for computing normative volume distributions and assessing patient-specific volumes? To answer these questions, we compared volumes of six brain regions calculated by five state-of-the-art segmentation methods: Erasmus MC (EMC), FreeSurfer (FS), geodesic information flows (GIF), multi-atlas label propagation with expectation–maximization (MALP-EM), and model-based brain segmentation (MBS). We applied the methods on 988 non-demented (ND) subjects and computed the correlation (PCC-v) and absolute agreement (ICC-v) on the volumes. For most regions, the PCC-v was good (), indicating that volume differences between methods in ND subjects are mainly due to systematic differences. The ICC-v was generally lower, especially for the smaller regions, indicating that it is essential that the same method is used to generate normative and patient data. To evaluate the impact on single-subject analysis, we also applied the methods to 42 patients with Alzheimer’s disease (AD). In the case where the normative distributions and the patient-specific volumes were calculated by the same method, the patient’s distance to the normative distribution was assessed with the z-score. We determined the diagnostic value of this z-score, which showed to be consistent across methods. The absolute agreement on the AD patients’ z-scores was high for regions of thalamus and putamen. This is encouraging as it indicates that the studied methods are interchangeable for these regions. For regions such as the hippocampus, amygdala, caudate nucleus and accumbens, and globus pallidus, not all method combinations showed a high ICC-z. Whether two methods are indeed interchangeable should be confirmed for the specific application and dataset of interest.

Segmentation of WM, GM, and CSF from the Brain MRIs Using an Advanced Metaheuristic Approach

Segmentation of brain tissues from the Magnetic Resonance Image (MRI) is vital to detect and diagnose brain-related diseases. Precise brain MRI segmentation is challenging because of the unpredictable anatomical structure of the brain, existence of Intensity Inhomogeneity (IIH), Partial Volume (PV) effects, and noise. Most of the brain disorders can be identified exactly by the segmentation of brain tissues, such as White Matter (WM), Gray Matter (GM), and Cerebrospinal Fluid (CSF). The optimization algorithmic approach is widely used for the brain image segmentation. In this work, a robust metaheuristic Multi-level Thresholding (MT) technique named as Advanced Wind Driven Optimization (AWDO) is employed for the brain MRI segmentation. The optimal MT values can be attained by combining the AWDO with the very popular Otsu’s objective function. The results obtained by the proposed AWDO in segmenting various brain images at an optimized value of threshold are compared with the groundtruths and have given the high values for the metrics such as sensitivity, specificity, segmentation accuracy, dice coefficient, and Precision. The values of the objective function, Peak Signal to Noise Ratio (PSNR), Standard Deviation (SD), and Convergence time of the proposed method are also being obtained at various threshold levels. The results are compared with the results of some existing algorithms like Multiplicative Intrinsic Component Optimization (MICO), Electromagnetism Optimization (EMO), Histogram based Darwinian Particle Swarm Optimization (HDPSO), Fully Convolutional Neural Networks (FCNN), and Advance Ant Colony Optimization (AACO). The results justified the notable performance of the proposed approach in segmenting different brain tissues.

Semi-Supervised Support Vector Machine for Digital Twins Based Brain Image Fusion.

The purpose is to explore the feature recognition, diagnosis, and forecasting performances of Semi-Supervised Support Vector Machines (S3VMs) for brain image fusion Digital Twins (DTs). Both unlabeled and labeled data are used regarding many unlabeled data in brain images, and semi supervised support vector machine (SVM) is proposed. Meantime, the AlexNet model is improved, and the brain images in real space are mapped to virtual space by using digital twins. Moreover, a diagnosis and prediction model of brain image fusion digital twins based on semi supervised SVM and improved AlexNet is constructed. Magnetic Resonance Imaging (MRI) data from the Brain Tumor Department of a Hospital are collected to test the performance of the constructed model through simulation experiments. Some state-of-art models are included for performance comparison: Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), AlexNet, and Multi-Layer Perceptron (MLP). Results demonstrate that the proposed model can provide a feature recognition and extraction accuracy of 92.52%, at least an improvement of 2.76% compared to other models. Its training lasts for about 100 s, and the test takes about 0.68 s. The Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) of the proposed model are 4.91 and 5.59%, respectively. Regarding the assessment indicators of brain image segmentation and fusion, the proposed model can provide a 79.55% Jaccard coefficient, a 90.43% Positive Predictive Value (PPV), a 73.09% Sensitivity, and a 75.58% Dice Similarity Coefficient (DSC), remarkably better than other models. Acceleration efficiency analysis suggests that the improved AlexNet model is suitable for processing massive brain image data with a higher speedup indicator. To sum up, the constructed model can provide high accuracy, good acceleration efficiency, and excellent segmentation and recognition performances while ensuring low errors, which can provide an experimental basis for brain image feature recognition and digital diagnosis.

An Effective Segmentation of Tissues from MR Brain Images

Segmentation of MR brain image is quite useful in detection of tissues and further diagnosis. However, precise segmentation of tissues plays a significant role in diagnosing the patient more effectively. Previously, there are plenty of approaches was implemented and however they were failed to detect the exact tissue which led to the failure diagnosis. Therefore, an accurate detection of tissue is required for effective diagnosis. Here, this article presented an efficient segmentation of MR brain image tissues. Our approach includes a hybrid clustering mechanism with pre-processed by median filter. In addition, tissue area also estimated for better diagnosis of patient. In terms of computational complexity and segmentation accuracy the superiority of proposed hybrid approach over conventional segmentation algorithms of simulation results was disclosed.

An Automatic MR Brain Image Segmentation Method Using a Multitask Quadratic Regularized Clustering Algorithm

In the diagnosis of clinical medicine, medical image processing plays a vital role and has become a hot issue in image processing. Magnetic resonance imaging not only provides convenience for treatment, but also brings help to the rehabilitation of patients. However, there are some unfavorable factors in MRI brain images, such as blurred boundary data, weak anti-noise ability, and so on. The classical fuzzy clustering algorithm has strong advantages, but the improved method is relatively simple, only adjusting the degree of membership or changing the distance algorithm to enhance the clustering effect. Therefore, this paper proposes a new multitask quadratic regularized clustering (MT-QRC) algorithm for MRI brain image segmentation, which improves the single-task clustering performance by transferring relevant knowledge between tasks. The proposed MT-QRC algorithm introduces the spatial information item controlled by the quadratic regularization term to replace the fuzzy index, which reduces the limitation of the fuzzy index in clustering and enhances the parameter flexibility.

MRI Brain Image Segmentation by Fully Convectional U-Net

When there is rapid growth in the research, and it will lead to use off large amount of data to get accurate results. When you are having large number of data then we require new techniques that will gives better performance in processing. The segmentation of a brain tumour is critical for both treatment and prevention. Various researchers proposed different neural network architectures to get better performance in segmentation of the brain tumour. processing this huge data is challenging and time taking process for computational and analysis. In this paper we are discussing about image segmentation by using fully conventional network U-Net. In the first stage we are performing some pre-processing on data sets by using adaptive filters. In the next step we are using U-Net architecture to perform segmentation and prediction of MRI brain images. In the next step we are performing Contrast Limited Adaptive Histogram Equalization (CLAHE) to equalize images. CLAHE takes care of over-amplification of the contrast. CLAHE operates on tiles of the image, rather than entire image.

Optimized Shannon and Fuzzy Entropy based Machine Learning Model for Brain MRI Image Segmentation

The pre-processing procedures for medical image segmentation are a crucial task in MRI image study. The medical image thresholding approaches are competent for bi level thresholding due to its' easiness, strength, fewer convergence period and accurateness. The efficiency can be maintained using an extensive search which can be employed for choosing the best thresholds. In this scenario, swarm intelligence-based learning algorithms can be suitable to gain the best thresholds. In this paper, we have focused in thresholding algorithm for segmentation of MRI brain image by maximizing fuzzy entropy and Shannon Entropy using machine learning and new evolutionary techniques. We have considered, Whale Optimization algorithm (WOA) in order to find the best outcome as well as compared the obtained results with the Shannon Entropy or fuzzy entropy-based examination that are fundamentally improved by Differential Evolution (DE), Particle Swarm Optimization (PSO), Social group optimization algorithm (SGO). It is discovered that overall operation could be effective by the strategy in features which can be captured through picture similarity matrix along with entropy values. We have observed that the proposed whale optimization model is able to better optimize the Shannon and fuzzy entropy compared to other swarm intelligence algorithms. It is also noticed that the new swarm intelligent algorithm i.e Social Group Optimization algorithm (SGO) is also performing better than the other two optimization algorithms i.e., Differential Evolution (DE), Particle Swarm Optimization (PSO) and providing very closer performance compared to Whale optimization algorithm. However, social group optimization algorithm requires little less CPU time than whale optimization algorithm.

Efficient Clustering of Unlabeled Brain DICOM Images based on similarity

Clustering has proven to be an effective method in the medical field for finding patterns in labelled and unlabelled datasets. This work is implemented over whole body CT scans (∼1TB) of 3500 patients in form of unlabelled DICOM images. The whole-body CT images have been anonymized for 30 attributes based on DICOM regulations and the Brain images alone are segmented using the DICOM tag element called ‘Protocol stack’. The segmented Brain images are efficiently grouped based on visual similarity using K-means clustering after performing feature extraction and dimensionality reduction. The results of the clustering can be furtherutilized by radiologists to perform labelling or find patterns in Brain CT scans of patients that are difficult where each scan consists of a varying number of slices during detection of Internal Bleeding. The efficiency of K-means is analyzed by performing computation over a different number of clusters (K) by applying silhouette scores to find optimal cluster.

Transductive Transfer Learning for Domain Adaptation in Brain Magnetic Resonance Image Segmentation.

Segmentation of brain images from Magnetic Resonance Images (MRI) is an indispensable step in clinical practice. Morphological changes of sub-cortical brain structures and quantification of brain lesions are considered biomarkers of neurological and neurodegenerative disorders and used for diagnosis, treatment planning, and monitoring disease progression. In recent years, deep learning methods showed an outstanding performance in medical image segmentation. However, these methods suffer from generalisability problem due to inter-centre and inter-scanner variabilities of the MRI images. The main objective of the study is to develop an automated deep learning segmentation approach that is accurate and robust to the variabilities in scanner and acquisition protocols. In this paper, we propose a transductive transfer learning approach for domain adaptation to reduce the domain-shift effect in brain MRI segmentation. The transductive scenario assumes that there are sets of images from two different domains: (1) source—images with manually annotated labels; and (2) target—images without expert annotations. Then, the network is jointly optimised integrating both source and target images into the transductive training process to segment the regions of interest and to minimise the domain-shift effect. We proposed to use a histogram loss in the feature level to carry out the latter optimisation problem. In order to demonstrate the benefit of the proposed approach, the method has been tested in two different brain MRI image segmentation problems using multi-centre and multi-scanner databases for: (1) sub-cortical brain structure segmentation; and (2) white matter hyperintensities segmentation. The experiments showed that the segmentation performance of a pre-trained model could be significantly improved by up to 10%. For the first segmentation problem it was possible to achieve a maximum improvement from 0.680 to 0.799 in average Dice Similarity Coefficient (DSC) metric and for the second problem the average DSC improved from 0.504 to 0.602. Moreover, the improvements after domain adaptation were on par or showed better performance compared to the commonly used traditional unsupervised segmentation methods (FIRST and LST), also achieving faster execution time. Taking this into account, this work presents one more step toward the practical implementation of deep learning algorithms into the clinical routine.

A machine learning classification approach based glioma brain tumor detection

AbstractThis article develops a computer aided fully automated method for detecting and classifying the glioma brain magnetic resonance imaging (MRI) using machine learning classification approach. The noise contents in source brain MRI image are detected and removed using ridgelet filter and then the edges in noise removed image are detected using fuzzy logic and then contrast adaptive local histogram equalization is applied on the edge detected brain image for enhancing the edge pixels. The Gabor transformation is applied on the enhanced brain image and the features are computed from this transformed image. The computed features are optimized using feature optimization technique genetic algorithm (GA) and the optimized features are classified using adaptive neurofuzzy inference system (ANFIS) classification method, which classifies the source brain MRI image into either glioma or non‐glioma brain image. Finally, fuzzy C means algorithm is applied on the glioma brain image to segment the tumor regions. The segmented tumor regions in glioma brain image is compared with manually tumor segmented brain image in order to evaluate the performance efficiency of the proposed system and the simulation results shows that the proposed works in this article achieves optimum performance with state of the art methods.

Detection of Brain Tumor in MR images using hybrid Fuzzy C-mean clustering with graph cut segmentation technique

In the medical image processing field segmentation task is very important process in order to diagnose any diseases in medical images taken from different machine. Medical images from MRI, CT scan, X-rays, Ultrasound and PET have different features, so segmentation process is very challenging task. In this paper, the graph theoretical approaches are proposed because it has flexibility representing any complex structure. Before segmentation process, MR images are pre-processed using Region of Interest, Inverse method and Boundary detection method. In this method the segmentation of MRI brain images is performed using Fuzzy C-mean clustering with graph cut techniques. FCM algorithm is proved to be efficient in term of computational rate by improving cluster center and modified way to selection of seed points. Seed point selection is by using FCM clustering derives new technique called Fuzzy C-mean seed selection (FCMSS). Image segmentation is performed using obtained path in the graph applied on set of cluster region of interest on MR images to detect brain tumors.

RETRACTED: An efficient approach for brain tumor detection and segmentation in MR brain images using random forest classifier

Nowadays, the most demanding and time consuming task in medical image processing is Brain tumor segmentation and detection. Magnetic Resonance Imaging (MRI) is employed for creating a picture of any part in a body. MRI provides a competent quick manner for analyzing tumor in the brain. This proposed framework contains different stages for classifying tumor like Preprocessing, Feature extraction, Classification, and Segmentation. Initially, T1-weighted magnetic resonance brain images are considered as an input for computational purpose. Median filter is proposed to optimize the skull stripping in MRI images. Abnormal brain tissues are extracted in low contrast, in addition to meticulous location of edges of affected tissue can be detected. Then, Discrete Wavelet Transform (DWT) and Histogram of Oriented Gradients (HOG) are performing feature extraction process. HOG is used for extracting the features like texture and shape. Then, Classification is performed through Machine learning categorization techniques via Random Forest Classifier (RFC), Support Vector Machine (SVM), and Decision Tree (DT). These classifiers classify the brain image as either normal or abnormal and the performance is analyzed by various parameters such as sensitivity, specificity and accuracy.

Changes in Brain Volume Associated with Trauma‐Focused Cognitive Behavioral Therapy Among Youth with Posttraumatic Stress Disorder

This study investigated group differences and longitudinal changes in brain volume before and after trauma-focused cognitive behavioral therapy (TF-CBT) in 20 unmedicated youth with maltreatment-related posttraumatic stress disorder (PTSD) and 20 non-trauma-exposed healthy control (HC) participants. We collected MRI scans of brain anatomy before and after 5 months of TF-CBT or the same time interval for the HC group. FreeSurfer software was used to segment brain images into 95 cortical and subcortical volumes, which were submitted to optimal scaling regression with lasso variable selection. The resulting model of group differences at baseline included larger right medial orbital frontal and left posterior cingulate corticies and smaller right midcingulate and right precuneus corticies in the PTSD relative to the HC group, R2 = .67. The model of group differences in pre- to posttreatment change included greater longitudinal changes in right rostral middle frontal, left pars triangularis, right entorhinal, and left cuneus corticies in the PTSD relative to the HC group, R2 = .69. Within the PTSD group, pre- to posttreatment symptom improvement was modeled by longitudinal decreases in the left posterior cingulate cortex, R2 = .45, and predicted by baseline measures of a smaller right isthmus (retrosplenial) cingulate and larger left caudate, R2 = .77. In sum, treatment was associated with longitudinal changes in brain regions that support executive functioning but not those that discriminated PTSD from HC participants at baseline. Additionally, results confirm a role for the posterior/retrosplenial cingulate as a correlate of PTSD symptom improvement and predictor of treatment outcome.

REVIEW ON AUTOMATIC PROCESSING OF BRAIN IMAGES FOR SEGMENTATION AND ABNORMALITY DETECTION

Brain tumour detection is very popular in the area of medical image processing. This is due to the sensitivity of brain functionality and inter structure. Any kind of ignorance towards the problems related with brain may cause serious impact on human life/life style. Therefore, early detection or diagnosis of abnormalities or tumours helps the doctors and patients to rectify the brain related health problems. The images are obtained through scanning techniques which are very common. Images obtained from the scanning needs to be segmented carefully for the future analysis and damage control procedures. In this paper, a detailed review on different types of segmentation techniques proposed by various authors is studied and compared for a clear understanding of existing segmentation techniques. They are tabulated to summarize different methodologies, segmentation techniques, and existing processes for further studies on Brain image segmentation. Finally, a brief understanding towards deep learning techniques is studied in this paper to understand their role in modern era for automated segmentation process

RETRACTED ARTICLE: Identification of lesion using an efficient hybrid algorithm for MRI brain image segmentation

Splitting the image into mutually exclusive regions is referred to as image segmentation. The demarcation, portrayal, and cognitive imaging of expanses of preference in any MR images is segmentation. Identifying the expanses of preference is considered the most complicated task since there are great diversities identified in the MRI images. Diversities in the image content, huddled objects, constriction, image noise, non-uniform object quality, and other factors are responsible for complications identified in the segmentation process. Various techniques and algorithms are being designed and made available for image segmentation. But still, there is a good scope for developing a resourceful, dissolute procedure for curative image segmentation. The proposed work aims to spot the disbelieving region from the related region in the MRI brain Image using Fractional Order BAT Algorithm Fuzzy C. In this paper, Delaunay triangulation (DT) recompenses and fractional order nature-inspired algorithms are used to segment the lesion's assembly from brain tissue meritoriously. The approach is corroborated with 450 MR images gained from different sources or subjects. The relevant features are extracted and based on the results of Fuzzy C Means, and the image is segmented as normal or abnormal. The method provided a considerable improvement over the existing mechanisms adopted for segmentation based on evaluation parameters like sensitivity, accuracy, Jaccard Index. The mechanism helps radiologists in validating their screening activities effectively.

LBTS-Net: A fast and accurate CNN model for brain tumour segmentation.

An accurate tumour segmentation in brain images is a complicated task due to the complext structure and irregular shape of the tumour. In this letter, our contribution is twofold: (1) a lightweight brain tumour segmentation network (LBTS‐Net) is proposed for a fast yet accurate brain tumour segmentation; (2) transfer learning is integrated within the LBTS‐Net to fine‐tune the network and achieve a robust tumour segmentation. To the best of knowledge, this work is amongst the first in the literature which proposes a lightweight and tailored convolution neural network for brain tumour segmentation. The proposed model is based on the VGG architecture in which the number of convolution filters is cut to half in the first layer and the depth‐wise convolution is employed to lighten the VGG‐16 and VGG‐19 networks. Also, the original pixel‐labels in the LBTS‐Net are replaced by the new tumour labels in order to form the classification layer. Experimental results on the BRATS2015 database and comparisons with the state‐of‐the‐art methods confirmed the robustness of the proposed method achieving a global accuracy and a Dice score of 98.11% and 91%, respectively, while being much more computationally efficient due to containing almost half the number of parameters as in the standard VGG network.

Spatial Rough Intuitionistic Fuzzy C-Means Clustering for MRI Segmentation

Medical image segmentation is the challenging problem in real-time applications due to the occurrence of noise and uncertainties between different tissues in the magnetic resonance images (MRI). To overcome the issue, spatial rough intuitionistic fuzzy C-means method has been proposed. The segmentation of MR brain image has been implemented by updating the MRI based on global spatial information of pixel with intuitionistic fuzzy c-means algorithm for segmenting cerebro spinal fluid, white matter and gray matter tissues. Intuitionistic fuzzy sets and rough sets have been used to deal with uncertainty and vagueness in medical images. Intuitionistic fuzzy sets are used for image representations by using non-membership value, hesitation along with the membership value for the MR image. The membership value and non-membership value have been obtained using fuzzy hexagonal membership and fuzzy complement function respectively. Further, roughness measures are done to determine the initial cluster centroids by considering lower and upper approximation and the fuzzy c-means clustering algorithm has been updates by the euclidean distance between the pixels based on global spatial information for segmenting MR brain image. The proposed method have been implemented and analysed with quantitative and qualitatively for the synthetic and real MR images. Experimental results exhibit a higher degree of segmentation accuracy on both synthetic and real MR images compared to existing methods and achieves better performance.

Automatic rat brain image segmentation using triple cascaded convolutional neural networks in a clinical PET/MR

The purpose of this work was to develop and evaluate a deep learning approach for automatic rat brain image segmentation of magnetic resonance imaging (MRI) images in a clinical PET/MR, providing a useful tool for analyzing studies of the pathology and progression of neurological disease and to validate new radiotracers and therapeutic agents. Rat brain PET/MR images (N = 56) were collected from a clinical PET/MR system using a dedicated small-animal imaging phased array coil. A segmentation method based on a triple cascaded convolutional neural network (CNN) was developed, where, for a rectangular region of interest covering the whole brain, the entire brain volume was outlined using a CNN, then the outlined brain was fed into the cascaded network to segment both the cerebellum and cerebrum, and finally the sub-cortical structures within the cerebrum including hippocampus, thalamus, striatum, lateral ventricles and prefrontal cortex were segmented out using the last cascaded CNN. The dice score coefficient (DSC) between manually drawn labels and predicted labels were used to quantitatively evaluate the segmentation accuracy. The proposed method achieved a mean DSC of 0.965, 0.927, 0.858, 0.594, 0.847, 0.674 and 0.838 for whole brain, cerebellum, hippocampus, lateral ventricles, striatum, prefrontal cortex and thalamus, respectively. Compared with the segmentation results reported in previous publications using atlas-based methods, the proposed method demonstrated improved performance in the whole brain and cerebellum segmentation. In conclusion, the proposed method achieved high accuracy for rat brain segmentation in MRI images from a clinical PET/MR and enabled the possibility of automatic rat brain image processing for small animal neurological research.

MR Brain Image Segmentation Using a Fuzzy Weighted Multiview Possibility Clustering Algorithm with Low-Rank Constraints

To expand the multiview clustering abilities of traditional PCM in increasingly complex MR brain image segmentation tasks, a fuzzy weighted multiview possibility clustering algorithm with low-rank constraints (LR-FW-MVPCM) is proposed. The LR-FW-MVPCM can effectively mine both the internal consistency and diversity of multiview data, which are two principles for constructing a multiview clustering algorithm. First, a kernel norm is introduced as a low-rank constraint of the fuzzy membership matrix among multiple perspectives. Second, to ensure the clustering accuracy of the algorithm, the view fuzzy weighted mechanism is introduced to the framework of possibility c-means clustering, and the weights of each view are adaptively allocated during the iterative optimization process. The segmentation results of different brain tissues based on the proposed algorithm and three other algorithms illustrate that the LR-FW-MVPCM algorithm can segment MR brain images much more effectively and ensure better segmentation performance.

WITHDRAWN: Multi-perspective scaling convolutional neural networks for high-resolution MRI brain image segmentation

Abstract The occurrence of defect over the soft tissues and nervous system is gradually increasing where Magnetic Resonance Imaging (MRI) is the most preferred method for performing the examination. The brain tumor MR image segmentation performs functionalities like image reconstruction of affected (diseased tissues) and qualitative analysis of infected and normal tissues. The image segmentation accuracy with the physician’s perspective relies over the shape, size, and location of lesions tissues, appropriate diagnostic strategies, and disease determination. The outcomes of this investigation rely over Multi-Perspective Scaling Convolutional Neural Networks (MPS-CNN) model for segmenting brain tumors more effectually and accurately. The multi-scale inputs are given to the proposed CNN model to overcome the necessity to select the appropriate input scale based on the tumor size, neighborhood tumor analysis based on scaled images, and adoption towards various tumor sizes. Therefore, the segmentation accuracy can be increased based on the input multi-scale brain tumor images. Also, the faster segmentation with multi-scaling process accelerates the speed of ensuring real-time segmentation process. This scaling process can effectively segment the brain images in the MRI which enhances the generalization process. It is utilized for predicting the brain lesion tissue of MRI. The simulation is carried out in MATLAB environment. The anticipated MPS-CNN is compared with prevailing approaches like CNN, FCN, U-Net, SegNet, Deep V3, and Deep FCN. And the MPS-CNN shows better trade-off in contrary to other approaches.