Synopsis Report on Leukemia Detection using Image Processing Under the Guidance of PDF

Preview

Summary

Synopsis Report on Leukemia Detection using Image Processing Submitted in partial fulfillment for the award of the degree Of In ELECTRONICS & TELECOMMUNICATION ENGINEERING By Meraj Kazi (25) Vishal Maurya (39) Faizan Shaikh (55) Deepak Upadhyay (66) Under the Guidance of Prof. Junaid Mandviwala ...

Description

Accelerat ing t he world's research.

Synopsis Report on Leukemia Detection using Image Processing Under the Guidance of Deepak Upadhyay

Related papers

Download a PDF Pack of t he best relat ed papers 

Ext ract ion of Nucleolus Candidat e Zone in Whit e Blood Cells of Peripheral Blood Smear Image… Ramin Solt anzadeh Whit e Blood Cell Analysis Using Wat ershed and Circular Hough Transform Technique Gdpindia pvt . Lt d Classificat ion of WBC for Leukemia det ect ion A Project Report Submit t ed in Part ial Fulfillment of Req… Gaurav Kumar

Synopsis Report on

Leukemia Detection using Image Processing Submitted in partial fulfillment for the award of the degree Of In

ELECTRONICS & TELECOMMUNICATION ENGINEERING By Meraj Kazi (25) Vishal Maurya (39) Faizan Shaikh (55) Deepak Upadhyay (66) Under the Guidance of Prof. Junaid Mandviwala

Department of Electronics and Telecommunication Engineering Rizvi College of Engineering Off Carter Road, Rizvi Complex, Bandra (West), Mumbai-50

University of Mumbai Year 2018-2019

I

Department of Electronic and Telecommunication Engineering

Rizvi College of Engineering, Off Carter Road, Bandra(W), Mumbai-400050

CERTIFICATE This is to certify that 1) Meraj Kazi (25) 2) Vishal Maurya (39) 3) Faizan Shaikh (55) 4) Deepak Upadhyay (66)

of B.E. Electronics and Telecommunication have successfully submitted synopsis report on “Leukemia Detection using Image Processing‟ in partial fulfillment of the Degree of Bachelor of Engineering in Electronics and Telecommunication under the guidance of ‘Prof. Junaid Mandviwala‟ from Rizvi College of Engineering, Bandra (West), Mumbai in the year 2018-19

Internal Guide Prof. Junaid Mandviwala

Internal Examiner

Head of Department Prof. R.S. Deshmukh

External Examiner

Principal Dr. Varsha Shah II

ABSTRACT For this project with the help of image processing technique we will detect Leukemia using microscopic image various image processing technique are used for identification of red blood and immature white cells. At the moment, identification of blood disorder is through visual inspection of microscopic images by examining changes like texture, geometry, colour and statistical analysis of images. Leukemia is one of the leading causes of death among human. Its cure rate and prognosis depends mainly on the early detection and diagnosis of the disease objective of this project will be to detect the leukemia affected cells and count it. According to detection of immature cells leukemia can be identified and also define that either it is chronic or acute. To detect immature cells number of methods are used like histogram equalization, linear contrast stretching, some morphological techniques like area opening, area closing, erosion, dilation. Watershed transform k-means, Histogram equalization and linear contrast stretching and share based features are accurate 72.2%, 72%, 73.7 and 97.8% respectively. For this project we will be using MATLAB software for detection of leukemia cells in the normal blood cells. This concept doesn‟t require any medicinal device or a person skilled in medicinal field. Require almost no man-power. This technology can come to use in detection many other disease like anemia, malaria, deficiency of vitamin B12, brain tumor detection etc.

I

CONTENTS

Chapter No.

Pg. No.

Topic

Chapter 1

Introduction

01

Chapter 2

Problem Definition & Objective

02

Chapter 3

Literature Survey

04

Chapter 4

Related Theory, Methodology and Algorithm

06

Chapter 5

Major Expected Results

15

Chapter 6

Tool to be Used

16

References

17

Table of Figures Sr. No

Figures

Page No

1.

Bone anatomy

3

2.

The Formation of Myeloid and Lymphoid

3

Series of Cell 3.

Block Diagram of Watershed Transform

7

4(a).

Original Image

8

4(b).

RGB to HSV color model

8

4(c).

Saturation component

8

II

4(d).

Derivation of gradient magnitude

8

4(e).

Area opening

8

4(f).

Opening-Closing

8

4(g).

Detect white cell using colored watershed

8

label matrix 4(h).

Detection of white cell superimposed

8

transparency on original image 5.

Block Diagram of K-Mean Segmentation

10

6(a).

Original Image

11

6(b).

Image labeled by Cluster Index

11

6(c).

Cluster of Red Cell

11

6(d).

Cluster of White cell

11

7(a).

Original Image

12

7(b).

RGB to gray image

12

7(c).

Linear contrast

12

7(d).

Histogram equalization image

12

7(e).

Addition of L and H

13

7(f).

Subtraction of L and H

13

7(g).

Addition of e and f

13

7(h).

Thresholding method

13

7(i).

Removal of small particles

13

7(j)`

Edge detection by sobel operator

13

Algorithm to count the cell

15

8.

III

Chapter 1 Introduction There are different types of white cells in our body. Leukemia is nothing but cancer of blood cells in which number of white cells is increasing and those are immature cells that destroy other cells. Today laboratory test takes longer interval of time to diagnose the disease of leukemia and it is also time consuming, prone to human error and also tedious. The ratio of white blood cell in our body is 1000:1. It means that 1 white blood cell is present between 1000 red cell. So if number of white blood cells increase remarkably in large number then the person is succumbed to suffer from the leukemia. It further falls into two type: acute and chronic. If number of white cell is increasing in our body then this immature cells start destroying another cells of our body. So we need to use the technology that identifies different types of blood cells within short duration of time in emergency. Furthermore it is vital to study in detail how to differentiate different cell and recognize it as immature cell and according to it, detect the leukemia. Acute and chronic also have two types. 1) Lymphocytic and 2) Myeloblastic that both are due to immature blast of lymphoid and myeloid cell respectively. The task can be improved and performed in near-real time environment using biomedical image processing.

1

Chapter 2 Problem Definition & Objective Leukemia is generated from the bone marrow. It can be cause of death if treatment is not started at right time. A thin material is situated inside each bone which is known as bone marrow. Every human body has mainly three type of blood cells:RBC(red blood cell), WBC(white blood cell), PLT(platelets).For this project main point of concern is to detect leukemia. So we are only concentrating on the count of WBC(leucocytes).There are two type of stem cell, myeloid and lymphoid stem cell. Myeloid stem cell emerges into myeloid blast. This myeloid blast is the reason for generating of RBC(erythrocytes),WBC(leucocytes) and platelets. Lymphoid stem cell also endues lymphoid blast which will generate only the white blood cell(WBC).Bone marrow produces abnormal white blood cells(WBCs).These abnormal cells should die after some time but in reality they do not die and they become numerous in count. The normal white blood cells are interrupted by those abnormal white blood cells in doing their normal work. And this type of situation is named as disease like “Leukemia”. Leukemia can be classified into Chronic and Acute leukemia. Chronic Leukemia:-Abnormal white blood cells perform like normal white blood cells and they will increase gradually. Acute Leukemia:-Abnormal white blood cells do not perform like normal cells and they will increase rapidly in number. The main objective was to enhance algorithms that can detect disease from human blood images during their earlier stages in order to prevent them from worsening. The separation of overlapping cells is a great help for the image quantitative analysis and image recognition. Using an algorithm based on watershed algorithm to divide the overlapping cells from each other, the results show good findings for the overlapping cells. With that, a software application that can separate overlapping cells in the blood images

for

better

detection

of

leukemia

was

developed. 2

Figure 1:Bone anatomy

Figure 2. The Formation of Myeloid and Lymphoid Series of Cell

3

Chapter 3 Literature Survey [1] Leukemia Detection using Digital Image Processing Techniques By Himali P. Vaghela, Hardik Modi, Manoj Pandya, M.B. Potdar There are different types of white cells in our body. Leukemia is nothing but cancer of blood cells in which number of white cells is increasing and those are immature cells that destroy other cells. Today laboratory test takes longer interval of time to diagnose the disease of leukemia and it is also time consuming, prone to human error and also tedious. The ratio of white blood cell in our body is 1000:1. It means that 1 white blood cell is present between 1000 red cell. So if number of white blood cells increase remarkably in large number then the person is succumbed to suffer from the leukemia. It further falls into two type: acute and chronic. If number of white cell is increasing in our body then this immature cells start destroying another cells of our body.

[2] Automatic Leukemia Detection Using Image Processing Technique This paper is about the proposal of automated leukemia detection approach. In a manual method trained physician count WBC to detect leukemia from the images taken from the microscope. This manual counting process is time taking and not that much accurate because it completely depends on the physician‟s skill. To overcome these drawbacks an automated technique of detecting leukemia is developed. This technique involves some filtering techniques and k-mean clustering approach for image preprocessing and segmentation purpose respectively. After that an automated counting algorithm is used to count WBC to detect leukemia. Some features like area, perimeter, mean, centroid, solidity, smoothness, skewness, energy, entropy, homogeneity, standard deviation etc. are extracted and calculated. After that neural network methodology is used to know directly whether the image has cancer effected cell or not. This proposed method has achieved an accuracy of 90%. 4

[3] Acute Lymphocytic Leukemia Detection by Image Processing Using MATLAB In the field of medicine, the blood cancer is found to be most hazardous as it affects the blood, bone marrow, lymph and lymphatic system. Among the three, most common types of blood cancer is leukemia and is found to be more prevalent. Leukemia is either acute or chronic. In persons with all, lymphoblasts are overproduced in the bone marrow and continuously multiply, causing damage and death by inhibiting the production of normal cells.. It is estimated that there will be 6,250 new cases of acute lymphocytic leukemia and an estimated 1,450 people will die of this disease in 2015. In the proposed methodology, the acute lymphocytic leukemia is detected by image processing technique using MATLAB. The electron microscope images are used as source images. The image processing involves contrast enhancement and segmentation for detecting the cancer cells in the blood and generating the results in a less span of time.

5

Chapter 4 Related Theory, Methodology and Algorithm Various methods have been applied to automate the task to find out leukemia cells and count it.

Watershed transform Lim Huey Nee et al. proposed methods for segmentation of white cells based on morphological operation, gradient magnitude and watershed transform. First image acquisition techniques is used then segmentation is done to separate the blast cell and back ground. For this method, first the RGB image is converted into HSV color model and saturated component is extracted for further processing and then find the gradient magnitude for the saturation component. This is used for edge detection. Moreover, sobel, canny, prewitt operators are used for the edge detection. After extracting the white cells from the image and elimination of the background and red cells, dilation or erosion process is carried out. Then watershed transform is carried out to separate the connected cell. Thus, leukemic cell can be identified and this method gives very accurate result. But the exact separation of cells cannot be done using this method.

6

IMAGE ACQUISITION P R

CONVERSION TO GREYSCALE

GRADIENT MAGNITUDE FORMATION

E P

WATERSHED TRANSFORM OF GRADIENT MAGNITUDE

R O C E

S HISTOGRAMIC EQUIVALENT

OPENING WATERSHED TRANSFORMED IMAGE RECONSTRUCTION

E G

S

M

S

E

I

OPENING WATERSHED TRANSFORMED IMAGE

N

N

T A

OPENING AND CLOSING AND CLOSING OF WATERSHED TRANSFORMED IMAGE

T I O

OPENING AND CLOSING OF WATERSHED TRANSFORMED IMAGE BY RECONSTRUCTION

N

DETECTION OF CANCER CELL

Figure 3: Block Diagram of Watershed Transform

7

Figure 4(a):Original Image

Figure 4(b):RGB to HSV color model

Figure 4(c):Saturation component

Figure 4(d):Derivation of gradient magnitude

Figure 4(e):Area opening

Figure 4(f):Opening-Closing

Figure 4(g):Detect white cell using colored watershed label matrix

Figure 4(h):Detection of white cell superimposed transparency on original image

8

Pre-Processing: Image pre-processing is a technique by means of which signal to noise ratio and image quality can be improved that will be helpful in further processing. The functions performed by preprocessing are listed Below 

Gray scale conversion



Contrast Enhancement

Conversion to Gray Scale: A grayscale image is supposed to contain only „Gray‟ color where the red, green and blue color components are said to have same intensity values and so processing becomes flexible when we specify only a single intensity value for each pixel, instead of taking three intensity values needed to be specified for each pixel in a color image. Microscopic images are found to possess the primary colors (RGB).

Gradient Magnitude Formation: Gradient depicts characteristics relating to the property of an object. By means of using the Sobel edge masks, imfilter and some simple arithmetic operation the gradient magnitude is formed. The gradient will be high at the borders of the objects and low inside the objects as in Fig. 8. Watershed Transform: The term watershed refers to a ridge that divides areas based on different pixel intensities. By employing watershed transform the gradient image is converted into RGB image with unique labeling based on intensity values.

9

K Means Clustering Technique To identify the abnormalities in blood cells or to identify the lymphoblast, the method of clustering techniques is used. After pre-processing of image feature extraction is done that gives useful information about image. The pre-processing techniques are only used for image enhancement. It does not give any necessary information of image. So initially acquisition process is done and then contrast enhancement is necessary to see clear image. After that RGB image is converted to HSI color model and then K-means clustering technique is applied for segmentation. Median filter is used to remove the noise from image. After feature extraction, image is classified by clustering techniques. Feature extraction is used to identify the white cells or the lymphoblast from image.

Start

Number of cluster k

No Centroid No object move group?

Yes End

Distance object to centroids

Grouping based on minimum distance

Figure 5: Block Diagram of K-Mean Segmentation

10

Figure 6(a): Original Image

Figure 6(b): Image labeled by Cluster Index

Figure 6(c): Cluster of Red Cell

Figure 6(d):Cluster of white cell

11

Histogram Equalization and Linear Contrast Stretching To perform this operation, image is first converted from RGB to gray level and for contrast enhancement, histogram equalization process is used. Then statistical parameter like mean and standard deviation is calculated and erosion or dilation technique is used for morphological operation. Here sobel operator is used to detect the edge of cells.

Figure 7(a):Original Image

Figure 7(c): Linear contrast

Figure 7(b):RGB to gray image

Figure 7(d):Histogram equalization image

12

Figure 7(e):Addition of L and H

Figure 7(g): Addition of e and f

Figure 7(i):Removal of small particles

Figure 7(f):Subtraction of L and H

Figure 7(h): Thresholding method

Figure 7(j):Edge detection by sobel operator

13

Methodology After getting the knowledge about all these techniques, it is concluded that Shape based features is used to detect different shapes like circle, rectangle, ellipse, squares etc. Our blood cells also have different size and shapes. So to detect geometrical shape of cells, this shape based features are very useful method to detect different type of cells and their shapes. In figure below algorithm is given for performing different image processing operation on leukemia detected images. This algorithm is useful to detect number of overlapping and non-overlapping cells or to count red and white cells. First RGB image is converted into gray level to reduced dimension of image. After that thresholding method is used to convert the image into binary form for more accurate analysis Otsu‟s method is best for thresholding, After that some morphological operation like area opening, closing, erosion, dilation. Here, area opening is performed to remove connected component Dilation is techniques to add pixels to boundary of objects and erosion is used to remove the pixel on object boundaries. After detecting the boundary of object, hole filling operation is performed that is used to detect perfect cell.

14

Figure 8:Algorithm to count the cell

15

Chapter 5 Tools to be used MATLAB MATLAB (matrix laboratory) is a multi-paradigm numerical computing environment and proprietary programming language developed by MathWorks. MATLAB allows matrix manipulations, plotting of functions and data, implementation of algorithms, creation of user interfaces, and interfacing with progr...