Medical scan and X-ray, and lack side effects on

Medicalimage processing is one of the most challenging and emerging fields. Detectingbrain tumor is the recent challenge faced by medical imaging researchers, it’schallenging because of complex structure of brain. Brain tumor is abnormalgrowth of cancerous cells in the brain that can even cause death.

  This paper provides a comprehensive overview of multiple methods and techniquesused for brain tumor detection along with classification algorithms. These well-knownmethods use MRI scanned images for tumor detection purpose, thatprovides better results than CT scan, ultrasound and X-ray. Byusing MATLAB software and applying various image processing techniquesdiscussed i.e Image pre-processing, Image enhancement, segmentation and featureextraction, brain tumor can be detected from MRI images of the brain. Keywords: ImageProcessing, MRI, Brain Tumor, Extraction, Segmentation, Classification1.    IntroductionInrecent times, medical image processing is an emerging and demanding field,which helps surgeons and doctors in analysis and diagnosis of complex diseasessuch as cancer, kidney or bladder stones and brain tumor etc.  the diagnosis or detection of brain tumor hasvital importance because it is the most common brain disease that can causedeath 1. Brain tumor detection is challenging because of brain’s complex structure.

Best services for writing your paper according to Trustpilot

Premium Partner
From $18.00 per page
4,8 / 5
4,80
Writers Experience
4,80
Delivery
4,90
Support
4,70
Price
Recommended Service
From $13.90 per page
4,6 / 5
4,70
Writers Experience
4,70
Delivery
4,60
Support
4,60
Price
From $20.00 per page
4,5 / 5
4,80
Writers Experience
4,50
Delivery
4,40
Support
4,10
Price
* All Partners were chosen among 50+ writing services by our Customer Satisfaction Team

Brain is a central part of human body that contains many cells, that grows anddivide. Brain tumor is intracranial mass due to abnormal growth of cells in thebrain. The brain tumor can either be cancerous or benign. Primary brain tumorstarts from the brain whereas, secondary brain tumor expands into brain throughsome other body part.Medicalimaging plays a vital role in accurate diagnosis of brain tumors. Differenttechniques are used for analysis and detection of brain tumor such Magnetic ResonanceImaging (MRI), Computed tomography (CT) scan, Ultrasound and X-ray.

MRI ismostly and widely used technique in medical imaging because it provides betterresults i.e high quality images than CT scan and X-ray, and lack side effectson body tissues. Magnetic field and Radio frequency pulses are used in MRIimaging to scan and generate images of structure and organs of body. It helpsdoctors to perfectly visualize anatomic structures of brain by creating ahigh-resolution image. Detection of brain tumor at an early stage is a majorissue, By using MATLAB tool and image processing techniques, not only detectionbut classification of brain tumor is also possible.

2.   LiteratureSurveyBraintumor detection is very crucial task because of the complex structure of brain,most of the patents die because ofinaccurate detection 1. Various image processingtechniques are being used for accurate detection, identification andclassification of brain tumor. 2 proposes a tumor detection, identification andclassification method, that detects brain tumor using segmentation and featureextraction with the aid of pixel intensity. The technique consists ofconsecutive smoothing stages to remove noise and high frequency components, non-maximumsuppression and region of interest (ROI) detection through thresholding. It helpsto measure tumor position and presents a method to prevent the spreading oftumor.3 focuses on detection of braintumor from MRI images. This paper proposes use of region growing method thatdefine boundaries of brain tumor and provides precise segmentation andidentification of brain tumor.

In this paper salt & pepper noise is addedduring identification process and then filtered out by median filter and lastlytumor is located using segmentation.4 proposes a method based on neuralnetwork and discrete wavelet transform.Neural network is used   foridentification of brain tumor, the neural network is trained for selectedfeatures which are extracted from the image and tumor can be detected. Herefusion method is used for detection by using multimode scanning of images whichgives relatively good results, and discrete wavelet transformation is appliedat image to get coefficient values. A fully automatic segmentation process fortumor is proposed, in which the algorithm used, integrate the images first andall noise is removed and then transformed into a new image.5 This paper proposes a strategy for detection and extraction of brain tumorfrom MRI images of patient’s brain. The method consists of various noiseremoval functions, morphological operations and segmentation.

MATLAB softwareis used for detection and extraction from MRI scanned images. Two staged algorithmsis applied, at first stage preprocessing of images is done and segmentation isdone in second stage. Morphological operations are performed later. They state that the stage of tumoris based on the area of the tumor. So, for this, size of the tumor can becalculated by calculating the number of white pixels in tumor binary image.Brain Tumor can be classified according to its type.

Numerous other Image processing methodologies are also being used for detection of brain tumorthrough MRI.6 presents a simple strategy for detection, using image pre-processing,segmentation and extraction method. They proposed Histogram thresholding,K-means clustering and Fuzzy C-means & support Vector Machine (SVM)methods. The methods presented includes various pre-processing steps such asnoise removal and RGB to Gray conversion.Differentimage processing techniques are used for brain tumor detection, the techniquesare applied on MRI scanned images, techniques used mainly consists of following4 steps: Pre-processing, segmentation, extraction and classification.

7  Fig1 Imageprocessing stepsA.   Preprocessingand Enhancement of an Image Pre-processingand image enhancement is the first step of image processing, for accuratedetection, finer details of the image are enhanced and noise is removed Fig 2.Firstly, film artifacts are removed from MRI images (x-ray marks and labelsetc) then different filters are used according to requirement such as isotropicfilter is used for removal of background noise, weighted median filter toremove salt and pepper noise etc. weighted median filtering technique givesbetter result than median, spatial or adaptive filters  Fig2Pre-processing and enhancement 7 B.    Segmentation Imagesegmentation is a necessary and critical step in image analysis, it is employedto extract various features of image. It’s a process of separating an imageinto different regions having identical or similar properties such as color, graylevel, brightness or contrast etc.

it separates or divide an image into multiple specificregions, pixels in each region exhibit high similarity and pixels that liebetween the regions exhibit high contrast 7-8. In brain tumorsegmentation tumor tissues are separated from normal brain tissues. Then toimprove the quality of images and limit the risk of distinct regions fusion inthe segmentation phase an enhancement process is applied. Multiple techniquesand approaches for image segmentation are used such as Thresholding approach,edge approach, region approach and clustering etc 8 each of which haveseveral advantages and disadvantages, therefore, they are selected or usedbased on requirement. The main approaches to segmentation are as follows:a)     Thresholding ApproachItis the most commonly used segmentation method in which pixels are allocated tocategories according to the range in which pixel lies. It uses intensityhistogram and determine the intensity values i.

e threshold values. Therefore,the image is segmented based on threshold value. Eq 1  where’v’ is the gray value and ‘t’ is the threshold value. After thresholding, theimage is segmented into two values 0 and 1 and gray image is converted intobinary.

2b)    Edge approachInthis approach, edge filter is applied, edges are detected in an image which areassumed to represent object boundaries and help identify objects. Edgesare identified by rapid transition of intensity, after identification pixels are linked together torepresent a boundary.In this approach edge position is given by either first order derivative or byzero crossing in second order derivative. c)     Region approachThisapproach focuses on finding theobject region rather than edges, it’s based on an assumption that borderingpixels in one region that have same values. Pixels are compared with theneighbors to see if congruence criteria satisfy, if it does so, the pixel isset to belong to the cluster as one or more of its neighbor.Thereare two types of region approaches: Region growing and Region splitting. In Regiongrowing an initial point (seed point) is defined, all the other surroundingpixels having same intensity value as seed point are connected to the seedpoint.

Whereas, in region splitting approach no seed point is defined and theimage is divided into unconnected regions, which are connected later based onsome condition.Regionbased approach uses different clustering algorithms, such as:       i.           K-means algorithmIt is an unsupervised, iterative clusteringmethod, also known as hard clustering algorithm. It is a widely used method,that separates a given dataset into k number of datasets or clusters, where eachcluster is defined by its centroid, that is a point in which sum of distancefrom all the objects in that cluster is minimized, therefore, Purpose of thisalgorithm is to minimize the sum of distances of all the objects to theircluster centers and the objects are set to belong to certaincluster 8, hence called hard clustering. It is a fast algorithm and robust to implement but has a drawbackthat it may not successfully find overlapping clusters and may also fail tocluster non-liner datasets or noisy data.

      ii.           Fuzzy clusteringThisclustering is also known as soft clustering. As boundary of tumor tissue isirregular, this fuzzy clustering technique can be helpful in getting better results.

In this approach objects located on boundaries of clusters can be member ofmany clusters, Objects are classified into different groups such as the pixelvalue of an image can belong to many clusters unlike K-means, the objects arenot forced to belong to certain clusters. Comparison between K-means and Fuzzyis given in Table 1.    iii.           Genetic Algorithms(GA)It is an optimization technique,consisting of three main operators: Recombination, Mutation and Selectionoperation. It operates on population of strings, at the start number of solutionsor populations are available, the solution from one population is utilized toform a new population which is superior than the old one. If some conditionsatisfies this process is repeated. It can easily be implemented and can solvehigher non-linearities but its computational cost is a major drawback.     iv.

           Particle swarm optimization(PSO)Itis a population based search technique, it is initiated by randomly selectedpopulations (particles) each of which have individual fitness value, that canbe calculated by fitness function. In segmentation optimal cluster centers aredetermined by PSO.  Unlike GA, Itlacks recombination operator.

Comparison between GA and PSO is given in Table 2. C.  Feature ExtractionPreciseextraction of a tumor is a critical task because of brain’s complex structure 8.Shape, location of tumor, size and composition are some major parameters thatare considered for feature extraction.Classificationof tumor is then done based on the results obtained by feature extraction.

     1.    Comparison Fuzzy clustering K-means Algorithm It is an extended form of hard clustering It is known as hard clustering Objects may be linked with multiple clusters Objects are set to belong to certain clusters Cluster center is based on distance between data points Each cluster has a center point i.e centroid Used for analysis based on only distance between various input data points Used for analysis based on location and distance between various input data points Table 1: Comparison between Fuzzy clustering and K-means Algorithm  Generic Algorithm Particle swarm optimization Implements three operations: Recombination, mutation and selection. Lacks recombination operator.

Operations are not labeled like GA Initially a discrete technique suitable for combinatorial problems. Continuous technique, not well suitable for combinatorial problems. High computational cost and more parameters to adjust Fewer parameters to adjust, Therefore, easier to implement. Table 2:Comparison between Genetic Algorithm and PSO Fig 2:Methodologies  2.   ConclusionMRI imaging is veryhelpful for analysis, diagnosis, and treatment of brain tumor & providesbetter results than CT scan, ultrasound and X-ray.

It helps in brain tumordetection by segmentation, which is a critical step because wrongidentification may lead to severe consequences. This paper gives an overview ofseveral state of the art methodologies used for the detection of brain tumorsuch as Image pre-processing, enhancement and extraction, and also variousalgorithms used for classification.Differentimage processing techniques are used for brain tumor detection, the techniquesare applied on MRI scanned images, techniques used mainly consists of following4 steps: Pre-processing, segmentation, extraction and classification. 7  Fig1 Imageprocessing stepsA.   Preprocessingand Enhancement of an Image Pre-processingand image enhancement is the first step of image processing, for accuratedetection, finer details of the image are enhanced and noise is removed Fig 2.Firstly, film artifacts are removed from MRI images (x-ray marks and labelsetc) then different filters are used according to requirement such as isotropicfilter is used for removal of background noise, weighted median filter toremove salt and pepper noise etc. weighted median filtering technique givesbetter result than median, spatial or adaptive filters  Fig2Pre-processing and enhancement 7 B.    Segmentation Imagesegmentation is a necessary and critical step in image analysis, it is employedto extract various features of image.

It’s a process of separating an imageinto different regions having identical or similar properties such as color, graylevel, brightness or contrast etc. it separates or divide an image into multiple specificregions, pixels in each region exhibit high similarity and pixels that liebetween the regions exhibit high contrast 7-8. In brain tumorsegmentation tumor tissues are separated from normal brain tissues.

Then toimprove the quality of images and limit the risk of distinct regions fusion inthe segmentation phase an enhancement process is applied. Multiple techniquesand approaches for image segmentation are used such as Thresholding approach,edge approach, region approach and clustering etc 8 each of which haveseveral advantages and disadvantages, therefore, they are selected or usedbased on requirement. The main approaches to segmentation are as follows:a)     Thresholding ApproachItis the most commonly used segmentation method in which pixels are allocated tocategories according to the range in which pixel lies. It uses intensityhistogram and determine the intensity values i.e threshold values. Therefore,the image is segmented based on threshold value.

Eq 1  where’v’ is the gray value and ‘t’ is the threshold value. After thresholding, theimage is segmented into two values 0 and 1 and gray image is converted intobinary. 2b)    Edge approachInthis approach, edge filter is applied, edges are detected in an image which areassumed to represent object boundaries and help identify objects. Edgesare identified by rapid transition of intensity, after identification pixels are linked together torepresent a boundary.In this approach edge position is given by either first order derivative or byzero crossing in second order derivative. c)     Region approachThisapproach focuses on finding theobject region rather than edges, it’s based on an assumption that borderingpixels in one region that have same values. Pixels are compared with theneighbors to see if congruence criteria satisfy, if it does so, the pixel isset to belong to the cluster as one or more of its neighbor.Thereare two types of region approaches: Region growing and Region splitting.

In Regiongrowing an initial point (seed point) is defined, all the other surroundingpixels having same intensity value as seed point are connected to the seedpoint. Whereas, in region splitting approach no seed point is defined and theimage is divided into unconnected regions, which are connected later based onsome condition.Regionbased approach uses different clustering algorithms, such as:       i.           K-means algorithmIt is an unsupervised, iterative clusteringmethod, also known as hard clustering algorithm. It is a widely used method,that separates a given dataset into k number of datasets or clusters, where eachcluster is defined by its centroid, that is a point in which sum of distancefrom all the objects in that cluster is minimized, therefore, Purpose of thisalgorithm is to minimize the sum of distances of all the objects to theircluster centers and the objects are set to belong to certaincluster 8, hence called hard clustering. It is a fast algorithm and robust to implement but has a drawbackthat it may not successfully find overlapping clusters and may also fail tocluster non-liner datasets or noisy data.      ii.

           Fuzzy clusteringThisclustering is also known as soft clustering. As boundary of tumor tissue isirregular, this fuzzy clustering technique can be helpful in getting better results.In this approach objects located on boundaries of clusters can be member ofmany clusters, Objects are classified into different groups such as the pixelvalue of an image can belong to many clusters unlike K-means, the objects arenot forced to belong to certain clusters. Comparison between K-means and Fuzzyis given in Table 1.    iii.           Genetic Algorithms(GA)It is an optimization technique,consisting of three main operators: Recombination, Mutation and Selectionoperation. It operates on population of strings, at the start number of solutionsor populations are available, the solution from one population is utilized toform a new population which is superior than the old one.

If some conditionsatisfies this process is repeated. It can easily be implemented and can solvehigher non-linearities but its computational cost is a major drawback.     iv.

           Particle swarm optimization(PSO)Itis a population based search technique, it is initiated by randomly selectedpopulations (particles) each of which have individual fitness value, that canbe calculated by fitness function. In segmentation optimal cluster centers aredetermined by PSO.  Unlike GA, Itlacks recombination operator. Comparison between GA and PSO is given in Table 2. C.

  Feature ExtractionPreciseextraction of a tumor is a critical task because of brain’s complex structure 8.Shape, location of tumor, size and composition are some major parameters thatare considered for feature extraction.Classificationof tumor is then done based on the results obtained by feature extraction.     1.

    Comparison Fuzzy clustering K-means Algorithm It is an extended form of hard clustering It is known as hard clustering Objects may be linked with multiple clusters Objects are set to belong to certain clusters Cluster center is based on distance between data points Each cluster h a center point i.e centroid Used for analysis based on only distance between various input data points Used for analysis based on location and distance between various input data points Table 1: Comparison between Fuzzy clustering and K-means Algorithm  Generic Algorithm Particle swarm optimization Implements three operations: Recombination, mutation and selection. Lacks recombination operator. Operations are not labeled like GA Initially a discrete technique suitable for combinatorial problems. Continuous technique, not well suitable for combinatorial problems. High computational cost and more parameters to adjust Fewer parameters to adjust, Therefore, easier to implement.

Table 2:Comparison between Genetic Algorithm and PSO Fig 2:Methodologies  2.   ConclusionMRI imaging is veryhelpful for analysis, diagnosis, and treatment of brain tumor & providesbetter results than CT scan, ultrasound and X-ray. It helps in brain tumordetection by segmentation, which is a critical step because wrongidentification may lead to severe consequences. This paper gives an overview ofseveral state of the art methodologies used for the detection of brain tumorsuch as Image pre-processing, enhancement and extraction, and also variousalgorithms used for classification.