Optimization of Medical Images Using Gabor Filter

International Journal of Pure and Applied Mathematics Volume 118 No. 24 2018 ISSN: 1314-3395 (on-line version) url: http://www.acadpubl.eu/hub/ Special Issue http://www.acadpubl.eu/hub/

Optimization of Medical Images using Gabor Filter

Mr.Veda Narayanan and Pauline Sheeba Department of ETE Sathyabama Institute of Science and Technology (Deemed to be University) Chennai-19, Tamilnadu, India [email protected]

May 22, 2018

Abstract a new approach in medical science is to automate the medical image segmentation based on multi scale analysis and adaptive thresholding. The accurate identification of the any dangerous changes in the body plays an important role in medical diagnosis of many diseases. In contrast to the existing methods for computer aided diagnosis which are either window-based or tracking based, we propose a novel scheme which combines optimization and applications of gabor filter to identify abnormalities under various conditions such as identifying tumors , and various other conditions that may prove fatal if ignored due to human errors . Our method includes a multiscale analytical scheme based on Gabor filters and scale multiplication, and adaptive thresholding. The experimental results demonstrate the feasibility and effectiveness of the proposed algorithms which are good for detecting the lumps and tumors in the image captured accurately, with robustness to denoise and enhance the re- sponses at low contrast Key Words:medical image, multiscale analysis, adaptive thresholding .

1 International Journal of Pure and Applied Mathematics Special Issue

1 Introduction

Medical image processing is the most challenging and highly wanted field. Brain tumor detection in magnetic resonance imaging(MRI) has become an emerging field of medical image processing. Segmen- tation of images is one of the most difficult tasks holds an impotant position in image processing which determine the quality of the of the final result .Image segmentation is the process of dividing an :image into different regions .the aim of this paper is to provide a review on automated tool for brain tumor segmentation using MRI scanned image datasets .detection and extraction of tumor from MRI scan images of the brain is done by Matlab software. Under the existing medical conditions, in addition to surgery and radiation therapeutic methods, there are no more effective treatments, and the patients condition can be alleviated and controlled, but extremely difficult to cure, which will cause great bur- dens in both mentality and economy for the patients. In this sense, the treatment of cancer is a major social problem in both economic and financial aspects. A good solution to this problem will have important social and practical significance. Commonly used medical imaging methods are Computed To- mography (CT), Positron Emission Tomography (PET), CT / PET, Magnetic Resonance Imaging (MRI), and so on

2 EXISTING TECHNIQUES

The existing techniques are discussed here. Few existing techniques are discussed here. The techniques that are discussed are Otsus optimal thresholding, snake active contour, SIFT (Scale Invariant Feature Transform), PCA (Principal Component Analysis), SVD (Singular Value Decomposition) and watershed. A. Region growing : In this technique the images are partitioned by organizing the nearest pixel of similar kind. It starts with a pixel (initial seed) that having similar properties. Accordingly the neighbouring pixels based on homogeneity criteria are appended progressively to the seed. In splitting process ,region get divided into sub regions that do not satisfy a given homogeneity criteria. Spliting and merg- ing can be used together and its performance mostly depends on

2 International Journal of Pure and Applied Mathematics Special Issue

the selected homogeneity criterion. Without tuning homogeneity parameters, the seeded region growing technique is controlled by a number of initial seeds. If the number of regions was approximately known & used it to estimate the corresponding parameters of edge detection. B. Clustering The method of clustering organizes the objects into groups based on some feature, attribute and characteristic. Hence a cluster consists of groups of similar objects. There are two types of clustering, supervised and unsupervised. In supervised type clustering, cluster criteria are specified by the user. In unsupervised type, the cluster criteria are decided by the clustering system itself. C. Soft-Computing A self-organizing map (SOM) or self-organizing feature map is a type of artificial neural network for unsupervised learning. SOMs organize in training and mapping mode. Training process builds map using vector quantization process and mapping automatically classifies a new input vector. SOM map consists of neurons or nodes. Self organizing maps each of which are neurons associated with a weight vector map data input vectors and position in the map space. The self-organizing maps a higher dimensional input space to a lower dimensional map space. Energy, entropy, contrast, mean, median, variance, correlation, maximum and minimum intensity values used to provide clear description of tumor. D. System analysis : Existing technique : Nowadays, brain tumor has become one of the main cause for increasing• mortality among children and adults. Based on some researches, it has been found that the number of people suffering and dying from brain tumors has been increased to 300 per year during past few decades. Existing technique has been practiced to determine tumor patients• response to treatment since long time ago. The radiologist has made series of cross-sectional diameter measurements for indi- cator lesions purposes by using axial, incremental CT image data. Later, these measurements will be compared with the previous measurement scans. bullet Nevertheless, the measurement of lesion diameter does not represent the exact assessment of tumor size due to some fac-

3 International Journal of Pure and Applied Mathematics Special Issue

tors, such as: Irregular lesions The lesions that grow other than sphere shape may not adequately represented by diameters changes; Different measurement between inter-observer and intra-observer Referred to the image selection used for the measurement and the location of lesion boundary; Different levels of scanning results collected from various di- agnoses• The lesions may not be captured exactly at the same spot from one diagnosis to another. Hence, it affects the lesions image in which causes comparisons between examinations becoming more difficult.

3 PROPOSED TECHNIQUE

I. Pre-processing generally means removing noise and improving or altering image quality to suit a purpose. For this study, only commonly used enhancement and noise reduction techniques were implemented. II. The image enhancement that the study is interested in should yield the result of more prominent edges and a sharpened image, noise will be reduced thus reducing the blurring or salt paper eﬀect from the image that might produce errors.

Fig: ﬂowchart of proposed system INPUT MRI IMAGE: Image segmentation is one of the fun- damental approaches of digital image processing. During past few years, brain tumor segmentation in magnetic resonance imaging (MRI) has become a popular research area in the ﬁeld of medical imaging system. MRI is used in radiology for analysing internal structures and makes easy to extract the required region.

4 International Journal of Pure and Applied Mathematics Special Issue

GRAY SCALE IMAGE Gray scale imaging is sometimes called ”black and white”,but technically this is misnomer in true black and white ,also known as halftone, and the only possible shades are pure black and pure white gray shading in a halftone image is obtained by considering the images as a grid of black dots on white background (or vice versa ) and the sizes of the individuals dots determine the apparent lightness of the gray in their vicinity. The lightness of the gray is directly proportional to the number representing the brightness levels of the colors. Grayscale imaging can be collectively called as the as the ranges of shades of gray. Grayscale can be collectively called as the ranges of the shades of gray.MRI images are used in the GENETIC ALGORITHM A genetic algorithm is a search heuristic that is inspired by Charles Darwins theory of natural evolution. This algorithm re- flects the process of natural selection where the fittest individuals are selected for reproduction in order to produce offspring of the next generation. Five phases are considered in a genetic algorithm. 1. Initial population 2. Fitness function 3. Selection 4. Crossover 5. Mutation The process begins with a set of individuals which is called a Population. Each individual is a solution to the problem you want to solve. An individual is characterized by a set of parameters (variables) known as Genes. Genes are joined into a string to form a Chromosome (solution). In a genetic algorithm, the set of genes of an individual is represented using a string, in terms of an alphabet. Usually, binary values are used (string of 1s and 0s). We say that we encode the genes in a chromosome.

5 International Journal of Pure and Applied Mathematics Special Issue

Figure 5: population , chromosome and gene

Selection The idea of selection phase is to select the fittest individuals and let them pass their genes to the next generation. Two pairs of individuals (parents) are selected based on their fitness scores. Individuals with high fitness have more chance to be selected for reproduction. Crossover Crossover is the most significant phase in a genetic algorithm. For each pair of parents to be mated, a crossover point is chosen at random from within the genes. For example, consider the crossover point to be 3 as shown below.

FIGURE 4.2 Crossover point

Offspring are created by exchanging the genes of parents among themselves until the crossover point is reached. MUTATION In certain new offspring formed, some of their genes can be sub- jected to a mutation with a low random probability. This implies that some of the bits in the bit string can be flipped. Mutation occurs to maintain diversity within the population and prevent premature convergence. TERMINATION The algorithm terminates if the population has converged (does not produce offspring which are significantly different from the previous generation). Then it is said that the genetic algorithm has provided a set of solutions to our problem. ALGORITHM OF GENETIC ALGORITHM Step 1: [Start] Generate random population of n chromosomes (suitable solutions for the problem) Step 2: [Fitness] Evaluate the fitness f(x) of each chromosome x in the population Step 3: [New population] Create a new population by repeating following steps until the new population is complete

6 International Journal of Pure and Applied Mathematics Special Issue

Step 4: [Selection] Select two parent chromosomes from a population according to their fitness (the better fitness, the bigger chance to be selected) Step4 : [Crossover] With a crossover probability cross over the parents to form new offspring (children). If no crossover was per- formed, offspring is the exact copy of parents. Step 5: [Mutation] With a mutation probability mutate new offspring at each locus (position in chromosome). Step 6 :[Accepting] Place new offspring in the new population Step 7:[Replace] Use new generated population for a further run of the algorithm Step 8:[Test] If the end condition is satisfied, stop, and return the best solution in current population Step 9:[Loop] Go to step 2 Particle Swarm Optimization (PSO) Particle Swarm Optimization (PSO) was invented by Russell Eberhart and James Kennedy in 1995. Originally, these two started out developing computer software simulations of birds flocking around food sources, then later realized how well their algorithms worked on optimization problems.Particle Swarm Optimization might sound complicated, but it’s really a very simple algorithm. Over a number of iterations, a group of variables have their values adjusted closer to the member whose value is closest to the target at any given mo- ment. Imagine a flock of birds circling over an area where they can smell a hidden source of food. The one who is closest to the food chirps the loudest and the other birds swing around in his direction. If any of the other circling birds comes closer to the target than the first, it chirps louder and the others veer over toward him. This tightening pattern continues until one of the birds happens upon the food. It’s an algorithm that’s simple and easy to implement. The algorithm keeps track of three global variables: Target value or condition • Global best (gBest) value indicating which particle’s data is currently• closest to the Target Stopping value indicating when the algorithm should stop if the• Target isn’t found Each particle consists of: Data representing a possible solution • A Velocity value indicating how much the Data can be changed •

7 International Journal of Pure and Applied Mathematics Special Issue

A personal best (pBest) value indicating the closest the particle’s• Data has ever come to the Target

fig: flowchart of particle swarm optimization GABOR FILTER A Gabor filter responds to edges and texture changes. When we say that a filter responds to a particular feature, we mean that the filter has a distinguishing value at the spatial location of that feature (when were dealing with applying convolution kernels in spatial domain, that is. The same holds for other domains, such as frequency domains, as well). Ksize: is the size of the Gabor kernel. If ksize = (a, b), we then have a Gabor kernel of size a x b pixels. As with many other convolution kernels, ksize is preferably odd and the kernel is a square (just for the sake of uniformity). Sigma: is the standard deviation of the Gaussian function used in the Gabor filter. Theta : is the orientation of the normal to the parallel stripes of the Gabor function. Lambda : is the wavelength of the sinusoidal factor in the above equation. Gamma : is the spatial aspect ratio. Psi : is the phase offset. Ktype : indicates the type and range of values that each pixel in the Gabor kernel can hold. Ksize: On varying ksize, the size of the convolution kernel varies. In the code above we modify the parameter ksize, while keeping the

8 International Journal of Pure and Applied Mathematics Special Issue

kernel square and of an odd size. We observe that there is no effect of the size of the convolution kernel on the output image. This also implies that the convolution kernel is scale invariant, since scaling the kernels size is analogous to scaling the size of the image. Here are a few results with varying ksize. For all the following images, sigma = 4.0, theta = 0, lambd = 10.0, gamma = 0.5, psi = 0, and ktype = cv2.CV 32F (i.e., each pixel of the convolution kernel holds a weight which is a 32-bit floating point number). Sigma: This parameter controls the width of the Gaussian envelope used in the Gabor kernel. Here are a few results obtained by varying this parameter. Theta This is perhaps one of the most important parameters of the Gabor filter. This parameter decides what kind of features the filter responds to. For example, giving theta a value of zero means that the filter is responsive only to horizontal features only. So, in order to obtain features at various angles in an image, we divide the interval between 0 and 180 into 16 equal parts, and compute a Gabor kernel for each value of theta thus obtained. Lambda: Heres the variation with lambda (theta is set to zero)

4 RESULT AND CONCLUSIONS

4.1 Experimental Results The data is ﬁrst presented according to example pictures of the image processes involved in the system. As discussed in the previous design part, the image used is of grayscale MRI brain scans. First we acquire the images and pre-processes it under a median ﬁlter. These are some of the examples the images:

9 International Journal of Pure and Applied Mathematics Special Issue

FIGURE 4.1 output of the proposed system clearly shows the tumor highlighted

5 CONCLUSION

We can conclude that morphological operations technique have proved to be very helpful in various image extraction and ﬁlter- ing techniques. The morphological operators can change the struc- turing elements of the image according to their use. Some operators like open, erode, dilate, close ,bounding box, region crop, have proved to be helpful in extracting the brain tumor from brain MRI image. threshold segmentation was used to work on the desired region of the image. Thus applying image subtraction we can get the ﬁnal brain tumor image.

References

[1] Natarajan P,Krishnan N,Natasha Sandeep Kenke,Shairya Nancy,Bhuvanesh Pratap Singh,”Tumor Detection Using threshold Operation In MRI Brain Images”IEEE international conference on computional intelligence and computing research,2012.

[2] Navneet Kaur,Mamta Juneja,”Brain Tumor Detec- tion,Demarcation and Quantiﬁcation via MRI”,International Journal Of Computer Applications,IJCA vol87-no.18,February 2014.

[3] B.Jegatheeskumar,Dr.P.Subbaiah,Dr.P.Ramesh,”Various Techniques for Brain Tumor Identiﬁcation and segmentation Approach in MRI Images ”,International Journal Of Computer Technology and Applications,IJCTA,Vol-5,Sep-Oct-2014.

[4] Nailah Afshan,Shaima Qureshi,Syed Mujitba Hus- sian,”Comparitive Study Of Tumor Detection Algo- rithms”,2014 IEEE Conference On Medical Imaging and emerging Communication Algorithm,2014.

10 International Journal of Pure and Applied Mathematics Special Issue

[5] Rajesh Patil,Dr. A S Bhalchandra,”Brain Tumor Extrac- tion From MRI images using MATLAB”,International Jour- nal Of Electronics & Communication Of Engineering And Soft Computing,IJECSSE,vol-2 Issue-1,2014.

[6] Jacek M Zurada,Tolga Ensari,Ehsaan Hosseini Asl,Jan Chor- wski,”Non Negative Matrix Factorization And Its Applications To Pattern Analysis And Text Mining”,2013 Federal Con- ference On Computer Science And Information System,IEEE 2013.

[7] S Ray,”On A Theoritical Property Of The Bhattacharya Co- eﬃcient As A Feature Evaulation Criterion”,Pattern Recogni- tion Letters,North Holland,1989.

[8] Aman chandra kaushik,Vandana Sharma,”Brain Tumor Seg- mentation From MRI Images And Volume Calculation Of Tumor”,International Journal Of Pharmaceutical Science Invention,Vol.2,Isuue-7,July-2013.