© 2016 IJSRSET | Volume 2 | Issue 2 | Print ISSN : 2395-1990 | Online ISSN : 2394-4099
Themed Section: Engineering and Technology
Morphological Detection of Abnormal Cells in Blood Sample of Humans
Mythili. A, Sreeja. G, Thamzhamuthan. N
SNS College of Technology, Coimbatore, Tamil Nadu, India
ABSTRACT
Image processing techniques are widely used in the domain of medical sciences for detecting various
diseases, infections, tumours, cell abnormalities and various cancers. Detecting and curing a disease on
time is very important in the field of medicine for protecting and saving human life. Mostly in case of high
severity diseases where the mortality rates are more, the waiting time of patients for their reports such as
blood test, MRI is more. The time taken for generation of any of the test is from 1-3 days. The current
system used by the pathologists for identification of blood parameters is costly and the time involved in
generation of the reports is also more sometimes leading to loss of patient’s life. Also the pathological
tests are expensive, which are sometimes not affordable by the patient. This paper deals with an image
processing technique used for detecting the abnormalities of blood cells in less time. The proposed
technique also helps in identifying the dead cells in the blood samples using morphological techniques
through cell opening and closing which gives upto 89% computational accuracy.
Keywords: Blood Cells, Morphology, Fluorescent Dye.
I. INTRODUCTION
The blood consists of a suspension of special cells
in liquid called plasma. Blood consists of 55 %
plasma, and 45 % by cells called formed elements.
The blood performs a lot of important functions. By
means of the haemoglobin contained in the
erythrocytes, it carries oxygen to the tissues and
collects the carbon dioxide (CO2). The reports will
get generated after chemical treatments on blood
samples which requires more time which imposes a
Fluorescent dye. The dead cancer cells absorb the
dye and appear blue colour on the microscopic slide.
It is also costly as the instruments used for
identification of the blood parameters are costly
[1][2]. The patients suffer due to both these reasons
physically as well as mentally. The proposed
system gives accuracy than the watershed method.
The first step of diagnosis will be counting of
cancer cells from the blood samples and next step
will be bone marrow aspiration which is taken from
the pelvic bones or from other bones.
II. METHODS AND MATERIAL
2. Types of Blood Cells
There are two types of blood cells: (A) Normal blood
cells & (B) Abnormal blood cells.
2.1 Normal Blood Cells
The various normal blood cells are: Erythrocytes,
Leucocytes, and Thrombocytes.
2.1.1 Erythrocytes
The erythrocytes are the most numerous blood cells i.e.
about 4-6 millions/mm3. They are also called red cells.
In man and in all mammals, erythrocytes are devoid of a
nucleus and have the shape of a biconcave lens. In the
mother vertebrates (e.g. fishes, amphibians, reptilians
IJSRSET162221 | Received : 30 March 2016 | Accepted : 30 March 2016 | March-April 2016 [(2)2: 413-419]
413
and birds), they have a nucleus [2]. The red cells are rich
in haemoglobin, a protein able to bind in a faint manner
to oxygen. Hence, these cells are responsible for
providing oxygen to tissues and partly for recovering
carbon dioxide produced as waste. However, most CO2
is carried by plasma, in the form of soluble carbonates.
exposure to hyper osmotic solutions. Pathological forms
are associated with uremia [5]. Echinocytes contain
adequate hemoglobin and the spiny knobs are regularly
dispersed over the cell surface, unlike those of
acanthocytes.
2.2.3 Dacrocyte (Tear Drop Cells)
2.1.2 Leucocytes
Leukocytes, or white cells, are responsible for the
defence of the organism. In the blood, they are much
less numerous than red cells. The density of the
leukocytes in the blood is 5000-7000/mm Leukocytes
divide in two categories: granulocytes and lymphoid
cells or agranulocytes. The term granulocyte is due to
the presence of granules in the cytoplasm of these cells.
In the different types of granulocytes, the granules are
different and help us to distinguish them. In fact, these
granules have a different affinity towards neutral, acid or
basic stains and give the cytoplasm different colors
[3][6].
Teardrop shaped red blood cells are found in
myelofibrosis and other myeloproliferative disorders,
pernicious anemia, thalassemia, myeloid metaplasia, and
some hemolytic anemias.
2.2.4 Sickle Cells
Sickle cells are red blood cells that have become
crescent shaped. When a person with sickle cell anaemia
is exposed to dehydration, infection, or low oxygen
supply, their fragile red blood cells form liquid crystals
and assume a crescent shape causing red cell destruction
and thickening of the blood. Since the life span of the
red blood cell is shortened.
2.1.3 Thrombocytes
2.3 Basic Overview of Paper
The main function of platelets, or thrombocytes, is to
stop the loss of blood from wounds (hematostasis). To
this purpose, they aggregate and release factors which
promote the blood coagulation [4]. Among them, there
are the serotonins which reduce the diameter of lessoned
vessels and slow down the hematic flux, the fibrin which
trap cells and forms the clotting. Even if platelets appear
roundish in shape, they are not real cells. In the smears
stained by Giemsa, they have an intense purple color.
2.2 Abnormal Blood Cells
The various abnormal blood cells are: elliptocyte,
echinocyte, dacrocyte, and sickle cells.
2.2.1 Elliptocyte
Elliptocytes are red blood cells that are oval or cigar
shaped. They may be found in various anemias, but are
found in large amounts in hereditary elliptocytosis.
2.2.2 Echinocyte (Crenated Red Blood Cells)
The aim of this paper is to detect the abnormal blood
cells by image processing technique. First we perform
the pre-processing technique by converting gray level
image into binary image. This research work will
propose a system which will help in reducing the cost as
well as waiting time of the patients for availing the
pathological reports. The system will be working as
follows:- Once the patient’s blood sample is collected, it
will be processed immediately and using an high end
camera’s in microscope, the images can be captured and
using the image processing techniques, the different
values of the desired parameters can be calculated
immediately[6][7]. Use of parameter dependent image
processing technique will definitely reduce the cost
involved and also will save the time of generating the
reports. This will definitely help to reduce the mortality
rate in high risk diseases. Section III, IV and V gives a
basic block diagram, morphological operation to detect
abnormal blood cells using image processing technique
respectively. Section VI discusses the results obtained.
Concluding remarks are given in section VII.
Echinocytes are red blood cells with many blunt spicules,
resulting from faulty drying of the blood smear or from
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
414
3. Basic Block Diagram
3.2 Image Enhancement
The dead cancer cells absorb the dye and appear blue
colour on the microscopic slide. Fig. 1.3 shows the basic
block diagram for detecting the abnormal blood cells.
The first step of diagnosis will be counting of cancer
cells from the blood samples.
For better segmentation of the blood cells, the imported
image has to be enhanced. This improves the quality of
the image in terms of details. Acquired image is been
pre-processed by converting the RGB image to
greyscale image and resized to 256*256. By enhancing
it improves the image contrast intensification and
brightness characteristics, reduce its noise cleaning or
smoothing, content or sharpen and edge sharpening or
crispening its details [10].
The proposed system would be an initiative to generate
the blood test reports in minimum time and will be cost
effective. Fig 1 and Fig 2 shows the normal and
abnormal cells in the blood.
Figure 1. Normal Blood Cells
Figure 2. Abnormal Blood Cells
Figure 3. Block diagram of the detection of abnormal
cells
3.3 Threshold of grey levels
3.1 Image Acquisition
The digital microscope is interfaced to a computer and
the microscopic images are obtained as digital images.
The resolution of the digital image depends on the type
of digital microscope used.
The greyscale image in again converted to binary image
by setting out the grey threshold value. A result of
thresholding gives a binary image. This involves
selecting only the area of interest in the image. Here
only the blood cells are selected, because they are the
areas of interest. Segmentation can be used for object
recognition, occlusion boundary estimation within
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
415
motion or stereo systems, image compression, image
editing or image database look up [15].
3.4 Morphological operations
A morphological operation helps to detect exact shape of
object. The data should be represented as a boundary or
as a complete region. It refers to certain operations
where an object is hit with a structuring element and
thereby reduced to more revealing shape. To create a
structuring element specified by a shapes structures like
disk, disk shaped approximation are suitable for
computing cells. Disk shaped structuring element is
approximated by specified radius from the origin of cells.
Morphological operation includes erosion, dilation,
opening and closing. In this proposed work
morphological erosion and dilation had been applied to
this image to eliminate small unwanted pixel and image
smoothing [11][12][13][9]. The erosion operation
uniformly reduces the size of objects in reaction to their
background and dilation expands the size of the objects.
Likewise opening and closing also can be used which is
like erosion followed by dilation (opening) and dilation
followed by erosion (closing) applied on the image.
Opening used to smooth the contours of cells and
parasites and closing used to fill the holes and gaps.
3.4.2 Morphological opening
The basic effect of an opening is somewhat like erosion
in that it tends to remove some of the foreground pixels
from the edges of regions of foreground pixels [5][7].
However it is less destructive than erosion in general. As
with other morphological operators, the exact operation
is determined by a structuring element. Fig 2.2 shows
the example of morphological opening.
The effect of the operator is to preserve foreground
regions that have a similar shape to this structuring
element or that can completely contain the structuring
element, while eliminating all other regions of
foreground pixels. After morphological closing the
image is been complemented, so that the back ground
image turns black and foreground image turns white
giving a clear vision on the cells.
3.4.1 Morphological closing
Figure 5. Example of Morphological Opening
Closing is the dual of opening, i.e. closing the
foreground pixels with a particular structuring element,
is equivalent to closing the background with the same
element. Fig 4 shows an example of morphological
closing [8]. The effect of the operator is to preserve
background regions that have a similar shape to this
structuring element, while eliminating all other regions
of back ground pixels.
3.4.3 Region properties and area calculation
Region properties help to extract many of the relevant
shape features without too much trouble. Since our
region taken for reference is disk shape, a disk shaped
structuring element is taken. It is used to count the
number dead cells in the blood sample through counting
the number of the white spots in the image. Averaged
area is also calculated to know the mean of the dead
cells in the blood sample. Elapsed run time gives the
estimation of the computational processing time to show
the result.
III. RESULTS AND DISCUSSION
Figure 4. Example of Morphological closing
The proposed algorithm is tested on the blood cell
images as shown in Fig 1. The total number of
blood cells present in the image is counted using
proposed algorithm (implemented using MATLAB).
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
416
The elapsed run time is 6.714272 seconds. The
mean area of the dead cells in the blood samples is
812.4000.
4.1 Results of Image Acquisition
The acquired image is a RGB microscopic image
taken from biological slides by higher end cameras.
It is been resized to 256*256. Resized image is
taken for the pre-processing. Fig 6 is the acquired
image of abnormal blood cells. A fluorescent dye is
injected to the slide and left for 3-4 hours. The
abnormal cells absorb the dye and appear blue.
Figure 7. Pre-Processed image of abnormal cells
Figure 8. Binary image of abnormal cells
4.3 Morphological opening and closing
Figure 6. Acquisition of abnormal image
4.2 Results of Pre processing
The next step after the image been acquired is preprocessing. The RGB image is converted into
greyscale image and threshold is determined. Fig 7
is the pre-processed image and Fig 8 is the
threshold image. Depending upon the threshold the
image is again converted to binary image for further
processing.
Rather than using watershed algorithms,
morphological operations have major advantage.
By applying morphological closing and opening the
results are in Fig 9 and 10
Figure 9. Morphological Opening
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
417
Figure 10. Morphological closing
Figure 12. Boundary output image
4.4 Result of Morphology
Figure 13. Final output image
Figure 11. Resultant of Morphological operation
The resultant is been complemented in order to
view the cells clearly.




Number of dead cancer cells = 30
Mean Area = 812.4000
Elapsed time is 2.186050 seconds
Run Time = 2.1861
4.5 Boundary Image Output
IV. CONCLUSION
For the cells that is been complemented, boundaries
are detected to super impose the complemented
boundary detected image on to grey scale image.
This will give the number of dead cells present in
the experimental slide. Fig 12 shows the boundary
of the abnormal cells in the blood sample.
The paper proposes an image processing technique
for detecting and counting the abnormal blood cells.
The proposed method detects the abnormalities in
blood cell in very less time and efficiently. Based
on the morphological operations, the abnormalities
of the blood cell can be segregated. Accuracy is up
to 87% and elapsed time is less than the existing
system i.e., Run time= 6.7143. The watershed
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
418
characterization” Indian J Med Res 134, October
method has the disadvantage that even small living
2011, pp 538-546538.
cells is segmented and counted as dead cells
involving miscalculation and risk to the lab [9] Abdul Nasir, A. S., Mustafa, N., and Mohd Nasir,
N. F., “Application of Thresholding Technique in
assistant and the patient. Finally with the help of
Determining Ratio of Blood Cells for Leukemia
proposed method, one of the major kinds of
Detection”, Proceedings of the International
leukemia called Acute Lymphoblastic Leukemia
Conference on Man-Machine Systems (ICoMMS)
(ALL) was determined.
11 – 13 October 2009, Batu Ferringhi, Penang,
V. REFERENCES
[1] Gonzalez, "Digital Image Processing Using
MATLAB," 2010.
[2] Siddharth Sekhar Barpanda, “Use of Image
Processing Techniques to Automatically Diagnose
Sickle-Cell Anemia Present in Red Blood Cells
Smear”, 2010.
[3] Devesh D. Nawgaje1, Dr. Rajendra D. Kanphade2”
Implementation of Computational Intelligent
Techniques for Diagnosis of Cancer Using Digital
Signal Processor” International Journal of
Application or Innovation in Engineering &
Management (IJAIEM) Web Site: www.ijaiem.org
Email:
editor@ijaiem.org,
editorijaiem@gmail.com, vol. 2, no. 1, January
2013.
[4] Robert T. Krivacic*, Andras Ladanyi, Douglas N.
Curry, H. B. Hsieh, Peter Kuhn, Danielle E.
Bergsrud,” A rare-cell detector for cancer” July 20,
2004 , vol. 101 , no. 29, pp. 10501–10504.
[5] Nasrul Humaimi Mahmood and Muhammad Asraf
Mans” Red blood cells estimation using hough
transform technique” Signal & Image Processing :
An International Journal (SIPIJ) Vol.3, No.2, April
2012.
[6] Zainb Nayyar “Blood cell detection and counting”
International Journal of Applied Engineering
Research and Development (IJAERD)ISSN(P):
2250-1584; ISSN(E): 2278-9383Vol. 4, Issue 2,
Apr 2014, 35- 40© TJPRC Pvt. Ltd.
[7] Stine-Kathrein Kraeft, Rebecca Sutherland,Laura
Gravelin, Guan-Hong Hu,Louis H. Ferland, Paul
Richardson,Anthony Elias, and Lan Bo Chen2”
Detection and Analysis of Cancer Cells in Blood
and Bone Marrow Using a Rare Event Imaging
System1” Vol. 6, 434–442, February 2000.
[8] Henri Wajcman and Kamran Moradkhani”
Abnormal
haemoglobins:
detection
&
MALAYSIA.
[10] S. Jagadeesh, E. Nagabhooshanam, and S.
Venkatachalam, “Image processing based approach
to cancer cell prediction in blood samples”,
International Journal of Technology and
Engineering Sciences, vol. 1, no. 1, ISSN: 23208007 2013.
[11] N., H., A., Halim, M., Y., Mashor, R.,
Hassan,“Automatic
Blasts
Counting
for
AcuteLeukemia
Based
on
Blood
Samples”,International Journal of Research And
Reviews in Computer Science, Vol. 2, No.
4,August 2011.
[12] Rupsa
Bhattacharjee,
Dr.Monisha
Chakroborty,”LPG-PCA Algorithm and Selective
Thresholding based automated Method: ALL and
AML Blast Cells Detection and Counting”
International Conference on Communications,
Devices and Intelligence System.2012.
[13] N., H., A., Halim, M., Y., Mashor, R., Hassan,
“Automatic Blasts Counting for Acute Leukemia
Based on Blood Samples”, International Journal of
Research and Reviews in Computer Science, Vol.
2, No. 4, August 2011, pp. 971-976.
[14] T. Markiewicz, S. Osowski, B., Marianska, L.,
Moszczynski, “Automatic Recognition of the
Blood Cells of Myelogenous Leukemia Using
SVM”, Proceedings of International Joint
Conference on Neural Networks, Montreal,
Canada, July 31 – August 4, 2005, pp. 2496-2501.
[15] Cseke, I., Geometry GIS System House Ltd.
Budapest 1025, Felso Zoldmali 128-1 30. Hungary
“A fastsegmentation scheme for white blood cell
images”, Proceedings of the 11th IAPR
International Conference on Pattern Recognition
Image, Speech and Signal Analysis, Aug. 30-Sept.
1992, Netherlands, pp: 530-533.
International Journal of Scientific Research in Science, Engineering and Technology (ijsrset.com)
419