Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI: [Essay Example], 1442 words GradesFixer

Haven't found the right essay?

Get an expert to write your essay!


Professional writers and researchers


Sources and citation are provided


3 hour delivery

This essay has been submitted by a student. This is not an example of the work written by professional essay writers.

Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI

Download Print

Pssst… we can write an original essay just for you.

Any subject. Any type of essay.

We’ll even meet a 3-hour deadline.

Get your price

121 writers online

Download PDF

To improve diagnosis and treatment medical imaging analysis have an important role but Magnetic Resonence Inaging (MRI) analysis have a key role, which is used to extract most relevant information. In medical diagnosis brain tumor diagnosis is much important as Brain tumors are most common disese of brain. Main purpose is to get information for patient treatmrnt and follow up and we have some important informations related to anatomical structures and potential abnormal tissues of the brain. Modeling of pathological brains and and the construction of pathological brain graphicses needs brain tumor segmentations. With the passage of time and with advancements of techniques and technologies in medical imaging still a lot of challenges are faced by medical imaging community like reproduceable segmentation and characterization of annormalities with sccurcy. Reasons behind these challenges are variety in brain tumor shapes, locations and intensity of different types of tumors. It also have possibility that some surrondding cells affected and produce confusing results in MRI , such possibilities are associated with edema and neorosis which is responsible for such changes. Exixting methods for brain MRI segmentations have space to improvise like improvements in automation, accuracy and applicability. This essay deals with the study of existing methods of detection and segmentation of brain tumor by using MRI.

Conventionally, in procedure of diagnosis, treatment and followup of the patient simple morphological technies are used on every image to segment the tissues or regions of intrest. Unfotunately such methods are unable to achieve all information available in an MRI. To use complete information with accurate results inaging analysis techniques are still at advanced development stage. A clear division in tumor detetions techniques is not possible as different methods are available by using multiple techniques, but we can divide them in three different groups,

  1. Region-based
  2. Counter based
  3. Fusion of region and boundry base methods.

Region-based methods search for out collections of voxels that share some portion of similarity. These methods decrease operator communication by automating some features of applying the low level operations, such as threshold selection, histogram analysis, classification, etc. They are may be supervised or non-Supervised.

Boundry based methods depends upom egde progress and internal forces like curvature with external forces like image gradiant which is helpful in explaining the boundry of brain structure. They are may be supervised or non-Supervised. Tumor segmentation methods are a mixture of region and boundry based methods. This class get succees to get attention of researchers because it uses information from two different classes, region and boundry based methods.

Region-based methods

In region based methods, algorithm searches for connection regions of pixels having some similar features like brightness and textures structures etc. only appling algorithm is not enough to solve detection and segmentation problem, some famous algorithms of region based methods includes thresholding, region growing, and classification.researchers develop sone new region based methods having mixture qualities which are helpful in detection and segmentation. New cata gories have : (1) classification-based; (2) clustering-based; (3) morphology-based; (4) graphics-based (5) prior knowledge-based; (6) texture-based; (7)feature extraction-based; (8) fusion-based; (9) neural network-based; (10) fuzzy-based; and (11) fractal-based method.

Still some time we become unable to provide clarity in these divisions for which below mention definations may play a key role. Few more details about methods are their requirements.

  1. Clustering-based techniques are those that use the fuzzy membership methods for segmenting brain tumors and are unsupervised classification.
  2. Mathematical morphology-based methods are those that use the features of mathematical morphology, such as structuring elements (SE) or masks or kernels as templates to convolve with the image, followed by binarisation using a given function or use gradient-dependent diffusion followed by linking.
  3. Graphics-based techniques perform the segmentation by registering and deforming the brain graphicses over patient images.
  4. Prior knowledge-based methods use prior knowledge of different structures, tissues and tumors of the brain to segment tumors.
  5. Texture-based techniques use statistical methods to compute textural features to distinguish brain tissues for segmentation of tumors.
  6. Feature extraction methods use a technique for extracting other features from patient images and they make a classification based on these features.
  7. Neural network-based methods are those that use an artificial neu- ral network (ANN) to learn classification parameters (using test MR data sets), this learned classification is then used to segment the patient images.
  8. Fusion-based methods use a fusion technique to combine the information of multimodality images (one device or multiple devices) for classification or clustering.
  9. Fuzzy methods are those that use fuzzy logic theory for segmentation. This class may has overlap with other classes and here we consider the fuzzy connectedness methods.
  10. Fractal-based methods use fractal concept to detect brain tumors in MR images.


Domain, classification algorithem can be supervised or un-superwised in image segmentation. Supervised classifier require inputs from user while Un-supervised classifier relies on cluster analysis. Below we have some supervised classifiers.

  • K-nearest neighbors (KNN),
  • Bayesian approach,
  • expectation maximization (EM),
  • Markov random field,
  • support vector machine (SVM).

Tumor segmentation using K-nearest neighbor (KNN)A simple classifier is the KNN classifier, in which each pixel is classified in the same class as the training data with the closest intensity. The KNN classifier is a nonparametric classifier as it makes no basic supposition about the statistical structure of the data. Vinitski et al. have developed a method which used patient-specific training to classify the T1-weighted, T2-weighted and PD-weighted images into 10 tissue classes. KNN classifier, that assigns labels to pixels based on the most regular label among the K closest training points under a reserve metric applied to the features. The KNN algorithm is effective method for multiclass classification, that is able to model nonlinear distributions. This is the first system which segments all components of the tumor (solid, edema and necrosis sections).

Disadvantages of the KNN algorithm include its dependent nature on the parameter K, large storage requirements, sensitivity to noise in the training data, and the undesirable behavior that can occur in cases where a class is under represented in the training data, which make it unsuitable for brain tumor segmentation in MRI.

Tumor segmentation using Bayesian approach

This is an supervised and parametric technique in which we assume that data will follow a multi variate, Guasian districution. [Corso et al., 2006] proposed a new process built on a combination of Bayesian model and graph based attraction for segmenting brain tumors. Four classes like non-data, brain matter, tumor and edema,are exhibited by Gaussian distributions with full covariance giving 9 parameters, which are cultured from the manually segmented data. Lastly a multilevel segmentation algorithm was introduced to segment the image, while in the first level each node is a pixel. This method is useful for segmentation of GBM tumors using T1-weighted, contrast enhanced T1-weighted and FLAIR images.Also it can be prolonged to vectorial variables to work on multi modality images. But it is comparatively slow and can only segment full-enhanced tumors like GBM. The other issue of this method is the displaying of the tumor by a Gaussian model, ehen the probability of tumor and edema do not always track Gaussian deliveries.

Tumor segmentation by expectation maximization (EM) was the first group use the EM algorithm to brain tumor segmentation. This method was developed on the basis of the work of [Leemput et al., 2001] for normal brain segmentation. The previous chances for the normal tissue classes are definite by the recorded spatial graphics to the patient images and the tumor spatial prior is calculated from the T1- weighted and contrast enhanced T1-weighted difference image. In this method the irregular pixels are measured as outliers from three normal classes. This method has 3 steps: firstly we determine abnormality area, by means of the EM algorithm and that the spatial priors for WM, GM, CSF and non-brain classes are steady to a verified spatial illustrations to patient image. For abnormal class a portion having sum of white matter and gray matter chances of graphics used.

The priors boundary of the segmentation class as the segmentation output can not fluctuate significantly from the graphics.

Tumor segmentation based on Markov random fields

Markov random fields (MRF) model normally useful to some problems growing in image processing. MRFs are used in some of the whole thing to expand the outcomes of the EM segmentation. This method innovative the EM results using a MRF, shared a structure to edge limit by means of a multi layer MRF, and prevailing a way to discrete partial capacity pixels from tumor pixels by making adaptive spatial prior for pixels that are at the limits of normal constructions. Solomon et al. proposed a tumor segmentation process using MRF. Further MRF model included into the EM context to progress the segmentation procedure. This method segments full enhanced tumors which did not study other components like edema and necrosis.

Remember: This is just a sample from a fellow student.

Your time is important. Let us write you an essay from scratch

100% plagiarism free

Sources and citations are provided

Cite this Essay

To export a reference to this article please select a referencing style below:

GradesFixer. (2020). Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI. Retrived from
GradesFixer. "Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI." GradesFixer, 12 Apr. 2020,
GradesFixer, 2020. Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI. [online] Available at: <> [Accessed 13 July 2020].
GradesFixer. Analysis Of The Methods Of Detection And Segmentation Of Brain Tumor By Using MRI [Internet]. GradesFixer; 2020 [cited 2020 April 12]. Available from:
copy to clipboard

Sorry, copying is not allowed on our website. If you’d like this or any other sample, we’ll happily email it to you.

    By clicking “Send”, you agree to our Terms of service and Privacy statement. We will occasionally send you account related emails.


    Attention! this essay is not unique. You can get 100% plagiarism FREE essay in 30sec

    Recieve 100% plagiarism-Free paper just for 4.99$ on email
    get unique paper
    *Public papers are open and may contain not unique content
    download public sample

    Sorry, we cannot unicalize this essay. You can order Unique paper and our professionals Rewrite it for you



    Your essay sample has been sent.

    Want us to write one just for you? We can custom edit this essay into an original, 100% plagiarism free essay.

    thanks-icon Order now

    Hi there!

    Are you interested in getting a customized paper?

    Check it out!
    Having trouble finding the perfect essay? We’ve got you covered. Hire a writer uses cookies. By continuing we’ll assume you board with our cookie policy.