automatic leukemia detector using digital image processing 3
TRANSCRIPT
Abstract— Blood cancers such as leukemia, Hodgkin
lymphoma, non-Hodgkin lymphoma, myeloma and
myelodysplastic syndromes are cancers that originate in the bone
marrow or lymphatic tissues. They are considered to be related
cancers because they involve the uncontrolled growth of cells
with similar functions and origins. The diseases result from an
acquired genetic injury to the DNA of a single cell, which
becomes abnormal (malignant) and multiplies continuously. The
accumulation of malignant cells interferes with the body's
production of healthy blood cells. Around 1,20,603 people in
India die every and Every 4 minutes one person is diagnosed with
a blood cancer due to leukemia.
I. INTRODUCTION
Blood cancers such as leukemia, Hodgkin lymphoma, non-
Hodgkin lymphoma, myeloma and myelodysplastic
syndromes are cancers that originate in the bone marrow or
lymphatic tissues. They are considered to be related cancers
because they involve the uncontrolled growth of cells with
similar functions and origins. The diseases result from an
acquired genetic injury to the DNA of a single cell, which
becomes abnormal (malignant) and multiplies continuously.
The accumulation of malignant cells interferes with the body's
production of healthy blood cells. Around 1,20,603 people in
India die every and Every 4 minutes one person is diagnosed
with a blood cancer due to leukemia. Leukemia can be
diagnosed:
From a blood test to measure the number of blood cells and
look for any abnormal cells.Slides of blood sample are
prepared and observed under the microscope to detect
abnormal shaped cells such as kidney shape.Or the blood
sample is detected for presence of immature cells in bone
marrow.The Conventional Method Of Diagnosis Suffer From
Following Disadvantages:-
All imaging techniques have to be repeated in cases
of suspected fungal infection.
Higher cost.
Human error of Visual Perception.
Using recent leukemia detection technique may lead to
incorrect conclusions. This may result due to human errors in
observing the peripheral sections under microscope. This may
lead to improper diagnosis of disease. Further numbers of
steps are involved in detection.
The most efficient method to overcome all
this drawbacks of present cancer detection techniq
ues is use of artificial intelligence. Basically here we use
“image processing” followed by “fuzzy logic”. These AI
techniques removes the human errors in detection, number of
steps involved are also less. And more importantly it is cost
effective
1. LEUKEMIA
White blood cells are made in your bone marrow, which is the
soft spongy centre of your bones. White blood cells are
involved in your body's immune system, a defense system that
protects you from infections. Your bone marrow makes the
most basic type of cells (called stem cells), and they can
develop further into three types of cells:
white blood cells - protect your body from infection
red blood cells - carry oxygen around your body
platelets - important for normal blood clotting
White blood cells are involved in your
body's immune system, a defense system that protects you
from infections. There are two main types of white blood
cells: myeloid cells and lymphocytes.
Leukemia affects three times as many adults
as children. It's the most common form of childhood cancer,
but this is because childhood cancers are rare. In the UK,
almost 7,000 people are diagnosed with leukemia every year ,
Figure 1: normal and leukemic cell in bone marrow
In leukemia, some of the white blood cells don't grow
properly. Leukemia is a type of cancer that affects the white
blood cells. In leukemia, white blood cells become abnormal,
and divide and grow in an uncontrolled way. They stay in the
bone marrow and keep reproducing in an uncontrolled way.
These abnormal white blood cells fill up the bone marrow and
Automatic Leukemia Detector Using Digital Image Processing
Ankush Bagwale, Nikhil Rai
Leukemia Detection Using Image Processing
prevent it from making healthy white blood cells. This means
the body is less able to fight off infections.
The abnormal white blood cells also prevent
bone marrow from making enough red blood cells and
platelets. A lack of red blood cells leads to less oxygen being
delivered to the organs and tissues of your body. This is called
anemia, and it can make you feel tired and breathless. A lack
of platelets can lead to problems with the blood-clotting
system, and results in bleeding and bruising much more easily
than usual.
of TRANS-JOUR.DOC or cut and paste from another
document and use markup styles. The pull-down style menu is
at the left of the Formatting Toolbar at the top of your Word
window (for example, the style at this point in the document is
“Text”). Highlight a section that you want to designate with a
certain style, then select the appropriate name on the style
menu. The style will adjust your fonts and line spacing. Do
not change the font sizes or line spacing to squeeze more
text into a limited number of pages. Use italics for
emphasis; do not underline.
2. RECENT DETECTION TECHNIQUE
Figure2: (a) & (b) leukemia affected cells
Leukemia can be diagnosed from a blood test to measure
the number of blood cells and look for any abnormal
cells.
Slides of blood sample are prepared and observed under
the microscope to detect abnormal shaped cells such as
kidney shape.
Or the blood sample is detected for presence of immature
cells in bone marrow.
People with suspected leukemia are referred
to a specialist doctor, usually a hematologist (an expert in the
treatment of blood disorders). Radiotherapy, chemotherapy
and some other techniques are used for treatment of cancer.
Sometimes doctors carry out further tests, such as:
CT (computerized tomography) scans - these can show
enlarged lymph nodes
MRI (magnetic resonance imaging) scans - to locate the
position of tumors
ultrasound - can be done to see if your spleen has been
affected by anemia
tissue typing - if your doctor has advised you to have a
bone marrow transplant
Diagnosis, investigation, treatment and
follow-up for people with leukemia usually take place at
specialist centers in hospitals.
3. NEED OF IMAGE PROCESSING
Using recent leukemia detection technique
may lead to incorrect conclusions. This may result due to
human errors in observing the peripheral sections under
microscope. This may lead to improper diagnosis of disease.
Further numbers of steps are involved in detection.
The most efficient method to overcome all
this drawbacks of present cancer detection techniques is use of
artificial intelligence. Basically here we use “image
processing” followed by “fuzzy logic”. These AI techniques
removes the human errors in detection, number of steps
involved are also less. And more importantly it is cost
effective.
4. INTRODUCTION TO IMAGE PROCESSING
A technique in which the data from an
image are digitized and various mathematical operations are
applied to the data, generally with a digital computer, in order
to create an enhanced image that is more useful or pleasing to
a human observer, or to perform some of the interpretation and
recognition tasks usually performed by humans.
An image is usually interpreted as a 2-D
array of brightness values. To digitally process an image, it is
first necessary to reduce the image into a number that can be
manipulated by the computer. Each number representing the
brightness value of the image at the particular location is
called a picture element or pixel. It is this pixel that undergoes
various types of operation such as point operation, local
operation, global operation etc. these operations, singly or in
combination , are the means by which the image is enhanced,
restored or compressed.
The figure (3) shows the effect of image processing.
Figure 3: effect of image processing
5. FUNDAMENTAL STEPS IN IMAGE
PROCESSING
5.1 IMAGE ACQUISITION
Image processing for microscopy
application begins with fundamental techniques intended to
most accurately reproduce the information contained in the
microscopic sample. This might include adjusting the
Leukemia Detection Using Image Processing
brightness and contrast of the image, averaging images to
reduce image noise and correcting for illumination non-
uniformities. Such processing involves only basic arithmetic
operations between images (i.e. addition, subtraction,
multiplication and division). The vast majority of processing
done on microscope image is of this nature.
5.2 IMAGE ENHANCEMENT
Enhancement refers to accentuation or
sharpening of image features, such as contrast, boundaries,
edges, etc. The process of image enhancement, however, in no
way increases the information content of the image data. It
increases the dynamic range of chosen features with the final
aim of improving the image quality. This is visible in the
following figure (4).
Figure 4: image enhancement
5.3 IMAGE RESTORATION
Sometimes we receive noisy images which
are degraded by some degrading mechanism. One common
source of degradation is the optical lens system in a digital
camera which acquires visual information.
When the degradation is due to atmosphere
(foggy weather), defocused camera or a relative accelerated
motion between the object and the focal plane of lens, the
conventional techniques of enhancement are not suitable to get
the original objects from the image. This is done by various
mathematical models and is called as reconstruction or
restoration of image.
Figure 5: image restoration
5.4 IMAGE COMPRESSION
Image compression, the art and science of
reducing the amount of data required to represent an image, is
one of the most useful and commercially successful
technologies in field of digital image processing. Compression
refers to the process of reducing the amount of data required to
represent that given quantity of information. Figure (6) shows
image compression
Figure 6 : image compression
5.5 IMAGE SEGMENTATION
In computer vision, segmentation refers to
the process of partitioning a digital image into multiple
segments (sets of pixels) (Also known as super pixels). The
goal of segmentation is to simplify and/or change the
representation of an image into something that is more
meaningful and easier to analyze. Image segmentation is
typically used to locate objects and boundaries (lines, curves,
etc.) in images. More precisely, image segmentation is the
process of assigning a label to every pixel in an image such
that pixels with the same label share certain visual
characteristics.
The result of image segmentation (figure
(7)) is a set of segments that collectively cover the entire
image, or a set of contours extracted from the image (see edge
detection). Each of the pixels in a region is similar with
respect to some characteristic or computed property, such as
color, intensity, or texture. Adjacent regions are significantly
different with respect to the same characteristics.
Leukemia Detection Using Image Processing
Figure 7 : image segmentation
5.6 COLOR IMAGE PROCESSING
The color of images can be altered in a
variety of ways. Colors can be faded in and out, and tones can
be changed using curves or other tools. The color balance can
be improved. In addition, more complicated procedures such
as the mixing of color channels are possible using more
advanced graphics. Image editors have provisions to
simultaneously change the contrast of images and brighten or
darken the image. Underexposed images can often be
improved by using this feature. Recent advances have allowed
more intelligent exposure correction whereby only pixels
below a particular luminosity threshold are brightened,
thereby brightening underexposed shadows without affecting
the rest of the image..
6. APPLICATION IN LEUKEMIA DETECTION
The acute leukemia is a disease of the
leukocytes and their precursors. It is characterized by the
appearance of immature, abnormal cells in the bone marrow
and peripheral blood. The aspirated marrow is found to be
infiltrated by abnormal cells. The recognition of the blast cells
in the bone marrow of the patients suffering from
myelogenous leukemia is a very important step in the
recognition of the development stage of the illness and proper
treatment of the patients.
Image segmentation is a division of the
image into different regions, each having certain properties. In
a segmented image, the picture elements are no longer the
pixels, but connected set of pixels, all belonging to the same
region. The segmentation techniques are used to separate the
individual cells from the set of cells creating the image. The
recognition and separation of individual cells from the image
of the blood cells is a very difficult task, since different
regions are of little grey level variations and the borders of
individual cells are hardly visible. Fig.8 presents the
exemplary image of the blast cells of the bone marrow
containing different types of cells.
Figure 8: the exemplary image of the bone marrow smear of
the acute leukemia patient containing different blast cells.
The individual cells are close to each other
and the borders among them are not well defined. The task of
segmentation of the image is focused on the automatic
recognition and separation of each cell for further processing.
6.1 EDGE DETECTION
Edge detection highlights image contrast.
Detecting contrast, which is difference in intensity, can
emphasize the boundaries of features within an image, since
this is where image contrast occurs. This is how human vision
can perceive the perimeter of an object, since the object is of
different intensity to its surrounding. Essentially, the boundary
of an object is a step change in the intensity levels. The edge is
at the position of step change. An ideal edge has the properties
of the model shown in figure 9(a) an ideal edge according to
this model is a set of connected pixel, each of which is located
at step transition in gray level.
In practice, optics, sampling, and other
image acquisition imperfections yield edges that are blurred,
with degree of blurring being determined by factors such as
the quality of image acquisition system, the sampling rate and
illumination condition under which the image is acquired. As
a result, edges are more closely modeled as having a “ramp
like” profile, such as the one shown in figure 9(b)
Figure 9 : (a) ideal edge (b) Ramp edge
The slope of the ramp is inversely
proportional to the degree of blurring ill the edge. In this
model, we no longer have a thin (one pixel thick) path.
Instead, an edge point now is any point contained in the ramp,
and edge would then be a set of such points that are connected.
Leukemia Detection Using Image Processing
The "thickness" of the edge is determined by the length of the
ramp, as it transitions from an initial to a final gray level. This
length is determined by the slope, which, in turn, is
determined by the degree of blurring. This makes sense:
Blurred edges tend to be thick and sharp edges tend to be thin.
.
Figure 10 :(a) vertical edge (b) Gray level first and second
derivative Profile
Detection of edge involves the use of first
and second derivatives. Figure 10(a) shows a horizontal gray-
level profile of the edge between the two regions. This figure
also shows the first and second derivatives of the .gray-level
profile. The first derivative is positive at the points of
transition into and out, of the ramp as we move from left to
right along the profile; it is constant for points in the ramp;
and-is zero in areas of constant gray level.
The second derivative is positive at the
transition associated with the dark side of the edge, negative at
the transition-associated with the light side of the edge, and
zero along the ramp and in areas of constant gray level. The
signs of the derivatives in Fig. 10(b) would be reversed for an
edge that transitions from light to dark.
We conclude from these observations that
the magnitude of the, first derivative can be used to detect the
presence of an edge at a point in an image (i.e., to determine if
a point is on a ramp). Similarly, the sign of the second
derivative can be used to determine whether an edge pixel lies
on the dark or light side of an edge.
We note two additional properties of the
second derivative around an edge: (1) It produces two values
for every edge in an image (an undesirable feature); and (2) an
imaginary straight line joining the extreme positive and
negative values of the second derivative would cross zero near
the midpoint of the edge. This zero-crossing property of the
second derivative is quite useful for locating the centers of
thick edges. 6.2 GRADIENT OPERATORS
First-order derivatives of a digital image are
based on various approximations of the 2-D gradient. The
gradient of an image f (x. y) at location (x, y) is defined as the
vector
It is well known from vector analysis that
the gradient vector points in the direction of maximum rate of
change of f at coordinates (x, y).An important quantity in edge
detection is the magnitude of this vector, denoted f,where'
This quantity gives the maximum rate of
increase of f(x, y) per unit distance in the direction of f. It
is a common (although not strictly correct) practice to refer to
f also as the gradient. We will adhere to convention and
also use this term interchangeably, differentiating between the
vector and its magnitude only in cases in which confusion is
likely. The direction of the gradient vector also is an important
quantity. Let α (x, y) represent the direction angle of the
vector f at (x, y). Then, from vector analysis,
where the angle is measured with respect to the x-axis. The
direction of an edge at (x, y) is perpendicular to the direction
of the gradient vector at that point.
7.3 LAPLACIAN OPERATOR
Second derivative can be computed using
laplacian operator.
Digital approximation
Leukemia Detection Using Image Processing
1) 7.4 THRESHOLDING AND LINKING
Once we have computed a measure of edge
strength (typically the gradient magnitude), the next stage is to
apply a threshold, to decide whether edges are present or not
at an image point. The lower the threshold, the more edges
will be detected, and the result will be increasingly susceptible
to noise, and also to picking out irrelevant features from the
image. Conversely a high threshold may miss subtle edges, or
result in fragmented edges.
If the edge thresholding is applied to just the
gradient magnitude image, the resulting edges will in general
be thick and some type of edge thinning post-processing is
necessary. A commonly used approach to handle the problem
of appropriate thresholds for thresholding is by using
thresholding with hysteresis. This method uses multiple
thresholds to find edges. We begin by using the upper
threshold to find the start of an edge. Once we have a start
point, we then trace the path of the edge through the image
pixel by pixel, marking an edge whenever we are above the
lower threshold. We stop marking our edge only when the
value falls below our lower threshold. This approach makes
the assumption that edges are likely to be in continuous
curves, and allows us to follow a faint section of an edge we
have previously seen, without meaning that every noisy pixel
in the image is marked down as an edge.
8. PROCEDURE
Based on above discussed concepts of
segmentation, the segmentation of cancerous blood cell (i.e.
edge detection) can be implemented with the help of following
steps in matlab programming.
Step 1: Read Image
Read in 'cell.tif', which is an image of a cancer cell.
I = imread('cell.tif'),figure, imshow(I), title('original image');
Original image
Figure:11
Thus figure (11) shows the original image to be processed.
Step 2: Detect Entire Cell
Two cells are present in this image, but only
one cell can be seen in its entirety. We will detect this cell.
Another word for object detection is segmentation. The object
to be segmented differs greatly in contrast from the
background image. Changes in contrast can be detected by
operators that calculate the gradient of an image. One way to
calculate the gradient of an image is the Sobel operator, which
creates a binary mask using a user-specified threshold value.
A threshold value is determined using the gray thresh
function. To create the binary gradient mask, we use the edge
function.
BWs = edge(I, 'sobel', (graythresh(I) * .1));
Figure, imshow (BWs), title ('binary gradient mask');
Binary gradient mask
Figure: 12
The binary gradient mask shows lines of
high contrast in the image as can be seen in the figure (12).
These lines do not quite delineate the outline of the object of
interest. Compared to the original image, you can see gaps in
the lines surrounding the object in the gradient mask. These
linear gaps will disappear if the Sobel image is dilated using
linear structuring elements, which we can create with the strel
function.
se90 = strel ('line', 3, 90);
se0 = strel ('line', 3, 0);
Step 3: Dilate the Image
The binary gradient mask is dilated using
the vertical structuring element followed by the horizontal
Leukemia Detection Using Image Processing
structuring element. The imdilate function dilates the
image.the image now appears as shown in figure (13).
BWsdil = imdilate (BWs, [se90 se0]);
figure, imshow(BWsdil), title('dilated gradient mask');
Dilated gradient mask
Figure:13
Step 4: Fill Interior Gaps
The dilated gradient mask shows the outline
of the cell quite nicely, but there are still holes in the interior
of the cell. To fill these holes we use the imfill function and
get the image as shown in figure (14).
BWdfill = imfill(Bwsdil, „holes‟);
figure, imshow(Bwdfill);
title(„binary image with filled holes‟);
Binary image with filled holes
Figure:14
Step 5: Remove Connected Objects on Border
The cell of interest has been successfully
segmented, but it is not the only object that has been found.
Any objects that are connected to the border of the image can
be removed using the imclearborder function. The
connectivity in the imclearborder function was set to 4 to
remove diagonal connections.
BWnobord = imclearborder (BWdfill, 4);
figure, imshow(BWnobord), title('cleared border image');
Cleared border image
Figure:15
Step 6: Smoothen The Object
Finally in order to make the segmented object look natural,
smoothening of the image is done by eroding the image twice
with a diamond structuring element. The diamond structuring
element can be created using the strel function
seD = strel('diamond',1);
BWfinal = imerode (BWnobord,seD);
BWfinal= imerode(BWfinal,seD);figure, imshow(BWfinal),
title ('segmented image');
segmented image
Figure:16
Step 7:Final Segmented Image
An alternate method for displaying the
segmented object would be to place an outline around the
segmented cell. The outline is created by bwperim function
BWoutline=bwperim(BWfinal);
segout=I;segout(BWoutline)=225;
figure, imshow(segout),title(„outlined original image‟);
Thus we obtain the final image as shown below in figure(17).
Outlined original image
Figure:17
Leukemia Detection Using Image Processing
9. CONCLUSION Thus using image processing segmentation,
edge of the cancerous blood cell can be obtained. Matlab
programming is used efficiently involving the concepts of
image segmentation to give the required edge to be checked
for any abnormal shape which assures cancer in a patient. This
image in which edge of cancerous cell is obtained can be
processed further for exact detection of cancer by any AI
technique.
Acknowledgment We would like to thank Dr. BHARAT PATIL of All India
Institute Of Medical Sciences (AIIMS ) New Delhi,Had it not
been the inspiration and motivation provided by Dr. Bharat
patil this research would definitely not progressed up the way
it is now. We plan to finish the work with Dr.Bharat Patil so
that we could provode this technology to the citizens of india.
at the least cost we can provide.we would also like to thank
Dr.Roopesh Ranjan for the financial support they provided by
funding this project.
References [1]„Digital image processing‟ by Woods and Gonzalez.
[2]http://www.cancerbackup.org.uk/
[3]http://www.leukaemiacare.org.uk/
[4]Feature extraction and image processing‟ by Mark S. Nixon
and Albrto S. Aguado
[5]Wikipedia- The Free Encyclopedia