Optics & Laser Technology 54 (2013) 225–231

Contents lists available at SciVerse ScienceDirect

Optics & Laser Technology

journal homepage: www.elsevier.com/locate/optlastec

A combined spatial-spectral method for automated white blood

cells segmentation
Qingli Li a,n, Yiting Wang b, Hongying Liu a, Jianbiao Wang c, Fangmin Guo a
a
Key Laboratory of Polor Materials and Devices, East China Normal University, Shanghai 200241, China
b
Institutes for Advanced Interdisciplinary Research, East China Normal University, Shanghai 200062, China
c
Ruinjin Hospital, Shanghai 200021, China

art ic l e i nf o a b s t r a c t

Article history: To overcome the shortcomings in the traditional white blood cells (WBCs) identification methods based
Received 25 March 2013 on the color or gray images captured by light microscopy, a microscopy hyperspectral imaging system
Received in revised form was used to analyze the blood smears. The system was developed by coupling an acousto-optic tunable
7 May 2013
filter (AOTF) adapter to a microscopy and driven by a SPF Model AOTF controller, which can capture
Accepted 17 May 2013
hyperspectral images from 550 nm to 1000 nm with the spectral resolution 2–5 nm. Moreover,
Available online 26 June 2013
a combined spatial-spectral algorithm is proposed to segment the nuclei and cytoplasm of WBCs from
Keywords: the microscopy hyperspectral images. The proposed algorithm is based on the pixel-wise improved
Microscopy hyperspectral imaging spectral angle mapper (ISAM) segmentation, followed by the majority voting within the active contour
Blood cells
model regions. Experimental results show that the accuracy of the proposed algorithm is 91.06% (nuclei)
Image segmentation
and 85.59% (cytoplasm), respectively, which is higher than that of the spectral information divergence
(SID) algorithm because the new method can jointly use both the spectral and spatial information of
blood cells.
& 2013 Elsevier Ltd. All rights reserved.

1. Introduction image processing and pattern recognition approaches. This can be

Microscopic blood cells analysis is a powerful diagnostic tool
for many types of diseases. Generally, there are three major
cellular constituents of blood: the red cells (or the erythrocytes),
the platelets, and the white blood cells (or the leukocytes). Among
these kinds of cells, the white blood cells (WBCs) refer to those
nucleated cells with diameters ranging from 6 μm to 20 μm which
play a very important role in the immune system. These WBCs can
eliminate germs such as bacteria and viruses, and fight cancer cells
and other toxic substances [1]. Therefore, identification and
inspection of WBCs in peripheral blood smear can provide valuable
information for disease diagnosis, such as leukemia, blood cancer,
and other blood-related diseases, which makes it to be one of the
most salient steps in hematological procedures [2]. Traditionally,
the manually and semi-automated methods have been used by
investigators to recognize and count the leukocyte cells [3,4].
These methods are generally time consuming, tedious, and opera-
tor fatigue producing, which can introduce unavoidable subjective
bias during the profile selection by the hematologists.
Nowadays, the rapid progress of information technology pro-
motes the automatization of WBCs analysis based on the modern
n
Corresponding author. Tel.: +86 2154345199.
E-mail addresses: qlli@cs.ecnu.edu.cn, tsingtlili@163.com (Q. Li).
called the computer-aided automatic analysis method, which has
been used for automatic identification and measurement of WBCs
in histological blood cells images. The most important step of the
computer-aided automatic analysis method is blood cell segmen-
tation, which can directly influence the feasibility and reliability of
analytical results. In the past decades, researchers have proposed
different algorithms to segment the WBCs automatically from the
microscopy images of blood smears. Harms and his coworker
proposed a segmentation strategy that characteristic color differ-
ence thresholds for each nucleus and cytoplasm combined with
geometric operations, probability functions, and a cell model [5].
This is one of the prior studies on automatic blood cells segmenta-
tion. Then, Park and Keller present a water snake (snakes on the
watershed) algorithm and applied it to segment the WBCs in bone
marrow images [6]. Consequently, different kinds of algorithms
have been proposed to segment the WBCs, such as the active
contour models [7], the feature space clustering based algorithm
[8], the stepwise merging rules and gradient vector flow snake
method [9], the multilevel thresholding [10], and the neural
network model [11]. These studies have shown that the automated
WBCs segmentation methods are faster and more objective
than the manually and semi-automatic methods. However, these
methods are associated with some other problems. First, most of
these methods were based on the RGB color images captured
by traditional light microscope which only contain the spatial

0030-3992/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.optlastec.2013.05.022
226 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231
information of blood cells. Second, due to uneven staining, color
mixing often occurs which leads to the smooth variations of the
average luminance in some regions of the image. In addition, the
microscopy imaging conditions, smear thickness, and background
illumination may result in biases and changes in shape, color, and
scale. All these factors influenced the accuracy of the automatic
segmentation results and lead to the algorithms complicated.
Therefore, there are still some difficulties and challenging pro-
blems to segment WBCs accurately based on the traditional
microscopic images.
The microscopy hyperspectral imaging technology may offer a
solution for these problems. Hyperspectral imaging, also known as
imaging spectroscopy or spectral imaging, is a technology that can
provide a digital image with far more spectral (color) information
for each pixel than traditional color cameras. The concept of
hyperspectral imaging was originally defined by Goetz in the late
1980s and the hyperspectral imaging system was originally devel-
oped for the remote sensing of the earth [12]. The advantage of
hyperspectral imaging technique is that it can acquire a reflectance
spectrum for each pixel in the image, which can be used to classify
the surface cover materials that can hardly be identified by
traditional gray or color imaging methods. In recent years, this
technology has been extended from the remote sensing field to the
life science field. Different kinds of microscopy hyperspectral
imaging system have been developed and used for biomedical
analysis of various biological tissues. For example, Morris et al.
have developed imaging spectrometers for fluorescence and
Raman microscopic analysis on rat brainstem [13]. Shonat et al.
used this technology to generate in vivo hemoglobin saturation
(SO2) and oxygen tension (PO2) maps in the cerebral cortex of mice
[14]. Some more recent studies have developed various micro-
scopy hyperspectral imaging systems based on prism grating
prism (PGP), acousto-optic tunable filters (AOTF), liquid crystal
tunable filter (LCTF), and variable interference filter (VIF) for
different biomedical applications [15,16]. All these studies shown
that the microscopy hyperspectral imaging technology can obtain
both images (structural information) and spectra (biochemical
information) of biological tissues which have the significant
advantages in the area of life science. However, only a few detailed
investigations on automated WBCs segmentation method based
microscopy hyperspectral imaging technology have been reported.
In the past decades, researchers have proposed different
segmentation algorithms to identify targets from the hyperspec-
tral images. Kruse et al. proposed a spectral angle mapper (SAM)
algorithm to detect spectrally active targets in Airborne Visible
Infrared Imaging Spectrometer (AVIRIS) data [17]. This is one of the
pixel based methods which have been commonly used in hyper-
spectral images segmentation. Then, some algorithms were pro-
posed to segment and classify objects from the multispectral and
hyperspectral images, such as the modified phase correlation
(MPC) [18], the active contours and graph cuts [19], the Markov
random field (MRF) model-based algorithm [20], and the support
vector machines (SVM) [21]. Moreover, some spatial-spectral
methodologies which consider spatial information along with
spectral information to improve the segmentation accuracy of a
hyperspectral image also have been investigated recently [22,23].
These algorithms have been used to analyze the remote sensing
images acquired in fields under the vegetation scene, the urban
environment, and the AVIRIS Indian Pines data set. However, little
studies have been reported on the automated white blood cells
segmentation using both the spatial and spectral information
extracted from the microscopy hyperspectral images.
In this study, an AOTF based microscopy hyperspectral imaging
(MHSI) system was developed and used to capture hyperspectral
images of blood smears. A preprocessing method was used to
remove the influence of noises and artifacts computationally.
Then, an automatic spatial-spectral segmentation algorithm based
on the microscopy hyperspectral images of blood smears is
presented and used for WBCs segmentation. Unlike those tradi-
tional light microscopy based segmentation methods, the hyper-
spectral based algorithm can segment the WBCs using both the
spatial and spectral information of blood smears. The experiment
results show that the proposed method is effective for WBCs
segmentation.
2. Materials and methods
2.1. Blood smear preparation
The samples used in this study are peripheral blood smears (or
peripheral blood films) which are glass microscope slides coated
on one side with thin layer of fresh venous blood. After collecting
the blood specimen, discard the first few drops of blood and then
make the smears as follow: place a drop of blood, about 2 mm in
diameter approximately an inch from the frosted area of the slide;
lightly balance another slide (spreader slide) on the fingertips and
place the spreader slide at a 301 angle in front of the drop of blood;
pull the spreader slide back toward the blood droplet and blood
spreads along the edge of the spreader slide; quickly push the
spreader across the slide with smooth action; allow the blood
smears to air-dry completely and finally dyed with Giemsa [24].
Those well-made peripheral smears which are thick at the frosted
end and become progressively thinner toward the opposite end
are selected and stored for future use.
2.2. Microscopy hyperspectral imaging system and image acquisition
The microscopy hyperspectral imaging system used in this
paper consists of a microscope (Nikon 80i, Nikon Co., Ltd., Japan),
an AOTF adapter (CVA200-0.55-1.0-L, Brimrose, USA), a SPF
Model AOTF controller (VFI-138.5-93-SPS-A-C2, Brimrose, USA),
a 1/1.8 in. high-density cooled charge coupled device detector
(CCD, DS-2MBWc, Nikon, Japan), a data collection and control
module (Camera Control Unit DS-U2, Nikon, Japan), and a personal
computer (A8800t, Lenovo, China). The front interface of the AOTF
adapter was coupled to the microscope with an F-mount adapter
and the back end was coupled to the CCD with C-mount. The light
source used in the microscope is a 120 W halogen lamp (Epi-
illumination, Eclipse 80i, Nikon). The AOTF adapter driven by the
SPF Model AOTF controller provides narrow bandwidth, rapid
wavelength selection, and intensity control. The minimum wave-
length selection sweep interval of the AOTF adapter is 20ns,
which makes it possible for the system to capture the microscopy
hyperspectral images quickly. The designed wavelength range of
the system is from 550 nm to 1000 nm and the spectral resolution
is 2–5 nm (2 nm at 543 nm; 5 nm at 792 nm).
The software running on the personal computer can control the
AOTF adapter to filter the light passed through the microscope and
record the image data on the CCD detector. In this way, the
microscopy hyperspectral imaging system can capture hundreds
of spectral bands covering the narrow spectral features of the
captured blood smears with high accuracy. As shown in Fig. 1,
the microscopy hyperspectral data can be visualized as a three
dimensional cube because of its intrinsic structure, where the cube
face is a function of the spatial coordinates and the depth is a
function of wavelength [25]. From the figure it can be seen that the
microscopy hyperspectral images of blood smears contain both
spectral and spatial information of blood cells, which makes it
possible to improve the accuracy of the automated WBCs segmen-
tation with some new hyperspectral based algorithms.
 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231
Fig. 1. The microscopy hyperspectral data cube.
2.3. Image preprocessing
The purpose of the preprocessing procedure is to remove
unwanted effects (such as noises) from the image, calibrate the
inhomogeneous spectral response of the system, and transform or
adjust the image as necessary for further processing. In the past
decades, researchers have proposed several kinds of preprocessing
methods for the hyperspectral images. In this study, the expecta-
tion maximization approach proposed by Wachman was used to
remove the effect of the acoustic angle distribution on the spectral
bandpass of an AOTF adapter [26]. Then, a calibration method
was used to remove the effects of the emission spectra of the
illumination sources, the transmission of the optics in microscope,
and the detection sensitivity of the CCD camera. In the calibration
process, a thin diffusing glass slide with approximately 95% trans-
mittance across the 550 nm to 1000 nm spectral range is selected
as the calibration specimen to capture the calibration data (both
the bright and dark images) using identical exposure settings
as the blood smears. Then the gray correction coefficient was
calculated to calibrate the spectral response of the system for each
scene of microscopy hyperspectral images of blood smears.
In addition, the microscopy hyperspectral images also need to be
normalized to get the transmittance data for further analysis.
The calibration and normalization procedure can be performed
according to the following equation:
DN white ðx; y; λÞ−DN dark ðx; y; λÞ
Tðx; y; λÞ ¼ log
ðDN blood ðx; y; λÞ−DN dark ðx; y; λÞÞT slide
where T(x, y; λ) denotes the normalized transmittance image of
blood smear, DNwhite(x, y; λ) denotes the intensity values of bright
calibration data, DNdark(x, y; λ) denotes the intensity values of
dark calibration data, DNblood(x, y; λ) denotes the intensity image
of blood smears, and Tslide denotes the transimittance of the
calibration slide specified by the manufacturer.
2.4. WBCs segmentation algorithm
Most automated WBCs segmentation method based on the
traditional RGB color blood cell images captured by light micro-
scopy can hardly get satisfied results because the color and the
intensity often change due to instability of staining, and light
variation and noise [2]. This makes more than 20% of all processed
blood samples still require microscopic review by experts [27].
Therefore, the multispectral and hyperspectral technology has
been used to extract the WBCs from a complicated background
and then segment them into their morphological components,
such as the nuclei and cytoplasm [28,29]. As the microscopy
227
hyperspectral images of blood smears can provide both the spatial
and the spectral information of blood cells, they make the WBCs
more distinguishable than those in the traditional RGB images.
However, the high dimensionality of microscopy hyperspectral
data (one scene often contains tens to hundreds of single band
images) usually limits the performances of the segmentation
techniques. The commonly used pixel-wise or spectral based
segmentation algorithms often lack information about spatial
structures of the image. In order to increase segmentation perfor-
mances, the spatial information is also needed to be integrated
into the segmentation process. In this paper, we proposed a
spatial-spectral algorithm to segment the WBCs by incorporating
the active contour models with the improved spectral angle
mapper (ISAM) algorithm.
The active contour models (ACM, or Snake) was first proposed
by Kass et al. as a technique for image segmentation, matching,
and tracking anatomic structures [30]. This method is based on the
utilization of deformable contours, which conform to various
object shapes and motions. An active contour can be considered
as a snake that bends its body into the edge of the object. It is a
planar curve defined by a set of connected points that iteratively
move from an initial position so as to minimize some formulation
of energy functional associated with the snake [31]. Representing
the position of the snake parametrically by v(s)¼ ¼ [x(s), y(s)],
where s is a normalized arc length, we can write its total energy
functional as follows:
Z 1 Z 1
Etotal ¼ EACM ðvðsÞÞ ds ¼ ½Eint ðvðsÞÞ þ Eext ðvðsÞÞ ds ð2Þ
0 0
  ∂vðsÞ2  ∂2 vðsÞ 2 
1    
Eint ¼ αðsÞ  þ βðsÞ  ð3Þ
2 ∂s ∂s2
Eext ¼ −γðsÞjΔIðvðsÞÞ2 j ð4Þ
where Eint represents the internal energy term which controls the
elastic properties of the snake, Eext represents the external energy
which comes from the image and is responsible for the deforma-
tion and movement of a snake toward application-specific image
features, and ΔI is the gradient of the image I.
One of the properties of this method is the ability to make
continuous edges even if the edge is weak or broken, which is very
suitable for WBCs segmentation. However, the active contour
models can hardly be performed directly on the microscopy
hyperspectral data as it generally contains hundreds of single
band images. To obtain a one-band image which would contain
ð1Þ enough information for active contour models, the principal
component analysis (PCA) method is used to map these single
band images onto a vector space where an actual ordering relation
exists. The PCA is a mathematical procedure that uses an ortho-
gonal transformation to convert a set of observations of possibly
correlated variables into a set of values of linear uncorrelated
variables called principal components. This transformation is
defined in such a way that the first principal component has the
largest possible variance. Then, the active contour models can be
performed on the first principal component to get the segmenta-
tion map.
To utilize the spectral information of microscopy hyperspectral
images for WBCs segmentation, the ISAM algorithm was intro-
duced to the spatial-spectral algorithm. The spectral angle mapper
(SAM) algorithm is one of the commonly used methods to
segment the target from the hyperspectral images in remote
sensing field. It was proposed by Kruse et al. to determine the
spectral similarity between two spectra by calculating the ‘angle’
between them [17]. Smaller angle between the test spectrum and
the reference spectrum represents closer matches to the target.
Therefore, the targets can be segmented from the hyperspectral
228 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231
images with a threshold value of SA. However, this method often
leads to some segmentation errors when the wavelengths shifting
several bands with the influence of noise. To overcome this
disadvantage, we use the improved spectral angle mapper (ISAM)
algorithm in this paper. The ISAM algorithm calculates the spectral
angle (SA) between the reference spectrum and the test spectrum
with shift forward and backward one or two bands according to
the following equation:
0 1
N 3. Results
B ∑ t iþj r i
!! ! ! B C
SAð T ; R Þ ¼ MaxðSAj ð T j ; R ÞÞ ¼ MaxB
B arccos s i¼1
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
@ N
∑ t 2iþj ∑ r 2i
i¼1
where j denotes the band number offset (j¼ ¼ −2, −1, 0, 1, and 2),
SAj denotes the spectral angle between the reference spectrum
and the test spectrum shifted j bands forward or backward. Then
the maximum one can be considered as the SA between the
reference spectrum and the test spectrum and compared with the
threshold. Therefore, the ISAM algorithm can get more accurate
results than the SAM algorithm as it can eliminate the influence of
wavelengths shift.
By combining the active contour models and the ISAM, we can
get the spatial-spectral algorithm to segment the WBCs from the
microscopy hyperspectral images of blood smears. Fig. 2 shows
the general flow-chart of the proposed segmentation algorithm.
At the input we have the N-band microscopy hyperspectral data
cube of blood smears. After preprocessing, the PCA transformation
is performed on the data and the first principal component
is selected for the active contour model analysis. Then, the ISAM
is performed on the microscopy hyperspectral images to segment
them into regions. Consequently, these independent candidate
segmentations results should be somehow combined into a single
final segmentation. Among all the combination methods, majority
vote (also named majority rule, decision fusion or label voting) is
by far the simplest for implementation. This method does not
need to assume prior knowledge of the behavior of the individual
segmentation, and it also does not require training on large
Fig. 2. Flow-chart of the proposed WBCs segmentation scheme.
quantities of representative recognition results from the experts
[32]. After the active contour models and the ISAM process, we use
the majority voting rule to fuse the segmentations by weighting
each candidate result equally and assigning to each pixel the label
that most segmentation results agree on. Therefore, the WBCs can
be segmented accurately by using both the spatial and spectral
information contained in the microscopy hyperspectral images.
C
ffi C
C ð5Þ
N A 3.1. Microscopy hyperspectral images of blood smears
i¼1
In this study, microscopy hyperspectral images of blood smears
were captured using the home made microscopy hyperspectral
imaging system with 100  immersion oil (numerical aperture
1.30) objective lens. For each blood smear, 80 single bands cover
the wavelength ranges from 550 nm to 1000 nm with the average
band width 5.6 nm were captured and stored in band sequential
format (BSQ). As each single band image contains 1024  1024 
12 bit pixels, one scene of microscopy hyperspectral images of
blood smear consists of approximately 160 Mb data. For the
convenience of processing, we resized the 160 Mb data cube
to four sub-cubes with the size of 512  512  80 (  12 bit) for
further analysis. A representative subtotal of microscopy hyper-
spectral images captured at various wavelengths is shown in Fig. 3.
From these single band images it can be seen that there are many
differences among different parts of the blood cells in different
single band images. These microscopy hyperspectral images can
provide both spatial and spectral information of blood cells which
is very useful for automatic WBCs segmentation.
3.2. Preprocessing result
After image acquisition, the microscopy hyperspectral images
can be calibrated and normalized by the methods proposed in
previous section. Fig. 4 shows some single band images extracted
from different scenes of microscopy hyperspectral images before
and after preprocessing. From the figure it can be seen that the
unwanted effects (noises and artifacts) contained in the original
images have been removed by the preprocessing procedure.
Therefore, the relative transmittance instead of the absolute
intensity data can be got for WBCs segmentation.
3.3. WBCs segmentation results
The automated nuclei and cytoplasm segmentation of WBCs
from the complex background is one of the prerequisite steps of
WBCs analysis. Most traditional segmentation algorithms are
based on the RGB color or gray images captured by light micro-
scopes which can only provide limited information of blood cells.
In this paper, we use the microscopy hyperspectral imaging
system to capture the images of blood cells and proposed the
spatial-spectral algorithm to segment the nuclei and cytoplasm of
WBCs. To experimentally check the method, 80 scenes of micro-
scopy hyperspectral images that contain various kinds of WBCs
were recorded. Then these images were preprocessed to remove
unwanted effects and segmented by the proposed algorithm.
To assess the performance of the WBCs segmentation algorithm,
single band images at wavelength 626 nm were selected from each
scene of microscopy hyperspectral data and manually segmented
by experienced pathologists using a graphic tool. The performance
of the proposed segmentation method was then compared with
that of the spectral information divergence (SID) algorithm, which
is a spectral based segmentation method proposed in our previous
study [29]. Fig. 5 shows the single band images at wavelength
 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231 229

Fig. 3. Single band images selected from microscopy hyperspectral data of a blood smear.

Fig. 4. Preprocessing results, (a)–(c) single band images before preprocessing, and (d)–(f) single band images after preprocessing.
230 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231

Fig. 5. WBCs segmentation results. The first column: single band images at wavelength 626 nm; the second column: segmentation results of SID algorithm; the third
column: segmentation results of the proposed algorithm; and the last column: ground truth labeled manually.

626 nm, the segmentation results of the SID algorithm and Table 1
the proposed algorithm, and the manually labeled images. From Comparisons of performance of the SID and the proposed algorithms.
the figure it can be seen that the spatial-spectral algorithm can
Measure SID (%) Proposed method (%)
get better results than the SID algorithm as both the spatial
and spectral information has been used to process each Nuclei Cytoplasm Nuclei Cytoplasm
pixel.
In order to compare the effectiveness of the proposed algo- NU 10.15 10.55 6.15 8.56
rithm and the SID algorithm quantitatively, the segmentation NO 4.77 12.46 2.79 5.85
Avg 85.08 76.99 91.06 85.59
results for each method were compared with the manually
extracted segmentation results (the ground truth) and the error
ratio estimated using the following equations [33]:
N U ¼ CardðM−ðM∩AÞÞ=SM ð6Þ
segmented region where a number closer to 1 represents a lower
error and higher accuracy, while a number closer to 0 represents a
N O ¼ CardðA−ðM∩AÞÞ=SA ð7Þ
higher error and lower reliability. Table 1 lists the performance
evaluation results of the SID and the proposed spatial-spectral
Avg ¼ ð1−ðN U ⊕N O ÞÞ ð8Þ
algorithm using Eqs. (6)–(8). From the table it can be seen that the
where M denotes the set of pixels in the manually segmented proposed method has a lower over- and under-segmentation error
region, A denotes the set of pixels in the automatically extracted ratio and a higher average accuracy Avg. Therefore, the proposed
region when using the SID algorithm or the proposed algorithm, segmentation method produced a better performance than the
SM and SA represent the total number of pixels in M and A, comparative algorithm as regards nuclei and cytoplasm. One of the
respectively, NU and NO denote the inaccuracy of the under reasons for the good performance of the proposed methods in
segmentation ratio and over segmentation ratio, respectively, the both NU and NO was that false positive regions were effectively
symbol Card(M) denotes the cardinality of set M, Avg represents removed by using both the spatial and spectral information of the
the accuracy between the ground truth and the automated microcopy hyperspectral images.
 Q. Li et al. / Optics & Laser Technology 54 (2013) 225–231
4. Conclusion
WBCs are the principal ingredient in the human immune
system. The morphometry and number of WBCs can be used as
a reference to identify the diseases because they vary with respect
to the age for each person [2]. However, manually and semi-
automated methods are time consuming works for biologists.
Therefore, some automated WBCs identification and segmentation
algorithms have been proposed to assist biologists. Most of these
methods are based on the color or gray images of blood smears
captured by light microscopy, which lead to these algorithms
complicated because of the limited information this kind of images
can provide. This paper presents a homemade microscopy hyer-
spectral imaging system to capture the blood smears and intro-
duces a preprocessing method to remove some effects such as the
inhomogeneous spectral response, noises and artifacts of the
system. Then a spatial-spectral scheme is proposed for segmenting
the nuclei and cytoplasm of WBCs. The new method is based on
the pixel-wise ISAM segmentation algorithm, followed by majority
voting within the ACM regions. Unlike those pixel-wise segmenta-
tion methods process each pixel independently according to
spectra without considering information about spatial structures,
the proposed method can improve the segmentation perfor-
mances by the incorporation of the spatial information into
the segmentation process. The experimental results show that
the combined spectral-spatial segmentation method can use both
the structure (spatial) and biochemical (spectral) information
of the blood smears to segment the WBCs, which makes the
segmentation algorithm more accurate than the RGB color based
methods.
As a preliminary research, we just presented the automated
WBCs segmentation algorithm based on the microscopy hyper-
spectral imaging technology and listed the corresponding results
to verify the possibility of this method. In the future study, we will
propose some new algorithms to perform the detailed analysis on
the blood smears, such as the classification of different kinds of
blood cells, the cell counting, the morphological parameters calcu-
lation, and the degree of oxygen saturation retrieval, which may be
useful for a better interpretation of further studies involving the
structural and biochemical changes and experimental models of
diseases.
Acknowledgment
This work is supported in part by the National Natural Science
Foundation of China (Grant no. 61177011, 61240006), the project
supported by the Shanghai Committee of Science and Technology,
China (Grant no. 11JC1403800), and the State Key Development
Program for Basic Research of China (Grant no. 2011CB932903).
References
[1] Khashman A. IBCIS: intelligent blood cell identification system. Progress in
Natural Science 2008;18:1309–14.
[2] Huang D-C, Hung K-D, Chan Y-K. A computer assisted method for leukocyte
nucleus segmentation and recognition in blood smear images. Journal of
Systems and Software 2012;85:2104–18.
[3] Dorini LB, Minetto R, Leite NJ. Semiautomatic white blood cell segmentation
based on multiscale analysis. Biomedical and Health Informatics, IEEE Journal
of 2013;17:250–6.
[4] Froom P, Havis R, Barak M. The rate of manual peripheral blood smear reviews
in outpatients,. Clinical Chemistry and Laboratory Medicine 2009;47:1401–5.
[5] Harms H, Aus HM, Haucke M, Gunzer U. Segmentation of stained blood cell
images measured at high scanning density with high magnification and high
numerical aperture optics. Cytometry 1986;7:522–31.
231
[6] Park J, Keller JM. Snakes on the watershed,. IEEE Transactions on Pattern
Analysis and Machine Intelligence 2001;23:1201–5.
[7] Yang L, Meer P, Foran DJ. Unsupervised segmentation based on robust
estimation and color active contour models. IEEE Transactions on Information
Technology in Biomedicine 2005;9:475–86.
[8] Jiang K, Liao QM, Xiong Y. A novel white blood cell segmentation scheme
based on feature space clustering. Soft Computing 2006;10:12–9.
[9] Ko BC, Gim J-W, Nam J-Y. Automatic white blood cell segmentation using
stepwise merging rules and gradient vector flow snake. Micron 2011;42:
695–705.
[10] Adollah R, Mashor MY, Rosline H, Harun NH. Multilevel thresholding as a
simple segmentation technique in acute leukemia images. Journal of Medical
Imaging and Health Informatics 2012;2:285–8.
[11] Khashman A. Investigation of different neural models for blood cell type
identification. Neural Computing & Applications 2012;21:1177–83.
[12] Goetz AFH, Vane G, Solomon JE, Rock BN. Imaging spectrometry for earth
remote sensing. Science 1985;228:1147–53.
[13] Morris HR, Hoyt CC, Treado PJ. Imaging spectrometers for fluorescence and
raman microscopy - acoustooptic and liquid-crystal tunable filters. Applied
Spectroscopy 1994;48:857–66.
[14] Shonat RD, Wachman ES, Niu WH, Koretsky AP, Farkas DL. Near-simultaneous
hemoglobin saturation and oxygen tension maps in mouse brain using an
AOTF microscope. Biophysical Journal 1997;73:1223–31.
[15] Li Q, Wang Y, Zhang J, Xu G, Xue Y. Quantitative analysis of protective effect of
erythropoietin on diabetic retinal cells using molecular hyperspectral imaging
technology. IEEE Transactions on Biomedical Engineering 2010;57:1699–706.
[16] Uhr JW, Huebschman ML, Frenkel EP, Lane NL, Ashfaq R, Liu HY, Rana DR,
Cheng L, Lin AT, Hughes GA, Zhang XJJ, Garner HR. Molecular profiling of
individual tumor cells by hyperspectral microscopic imaging. Translational
Research 2012;159:366–75.
[17] Kruse, F.A., Lefkoff, A.B., and Boardman, J.W., The spectral image processing
system (SIPS)-software for integrated analysis of AVIRIS data, presented at the
Summaries of the 4th Annual JPL Airborne Geoscience Workshop, Pasadena,
1992.
[18] Ertuerk A, Ertuerk S. Unsupervised segmentation of hyperspectral images
using modified phase correlation. IEEE Geoscience and Remote Sensing Letters
2006;3:527–31.
[19] De La Vega SH, Manian V. Object segmentation in hyperspectral images using
active contours and graph cuts. International Journal of Remote Sensing
2012;33:1246–63.
[20] Sarkar A, Vulimiri A, Paul S, Iqbal J, Banerjee A, Chatterjee R, Ray SS.
Unsupervised and supervised classification of hyperspectral imaging data
using projection pursuit and Markov random field segmentation. International
Journal of Remote Sensing 2012;33:5799–818.
[21] Mountrakis G, Im J, Ogole C. Support vector machines in remote sensing:
a review. ISPRS Journal of Photogrammetry and Remote Sensing 2011;66:
247–59.
[22] Gorretta N, Rabatel G, Fiorio C, Lelong C, Roger JM. An iterative hyperspectral
image segmentation method using a cross analysis of spectral and spatial
information. Chemometrics and Intelligent Laboratory Systems 2012;117:
213–23.
[23] Li J, Bioucas-Dias JM, Plaza A. Spectral-Spatial hyperspectral image segmenta-
tion using subspace multinomial logistic regression and Markov random
fields. IEEE Transactions on Geoscience and Remote Sensing 2012;50:809–23.
[24] Corrons JLV, Albarede S, Flandrin G, Heller S, Horvath K, Houwen B, Nordin G,
Sarkani E, Skitek M, Van Blerk M, Libeer JC, Eqalm. Guidelines for blood smear
preparation and staining procedure for setting up an external quality assess-
ment scheme for blood smear interpretation. Part I: control material. Clinical
Chemistry and Laboratory Medicine 2004;42:922–6.
[25] Rajpoot, K.M. and Rajpoot, N.M., Wavelet based segmentation of hyperspectral
colon tissue imagery, Proceedings INMIC 2003. 7th International Multi-Topic
Conference (IEEE Cat. no.03EX757), pp. 38–43|x+497, 2004.
[26] Wachman ES, Niu WH, Farkas DL. Imaging acousto-optic tunable filter with
0.35-micrometer spatial resolution. Applied Optics 1996;35:5220–6.
[27] Ceelie H, Dinkelaar RB, van Gelder W. Examination of peripheral blood films
using automated microscopy; evaluation of Diffmaster Octavia and Cellavision
DM96. Journal of Clinical Pathology 2007;60:72–9.
[28] Guo N, Zeng L, Wu Q. A method based on multispectral imaging technique for
white blood cell segmentation. Computers in Biology and Medicine 2007;37:
70–6.
[29] Guan Y, Li Q, Wang Y, Liu H, Zhu Z. Pathological leucocyte segmentation
algorithm based on hyperspectral imaging technique. Optical Engineering
2012;51.
[30] Kass M, Witkin A, Terzopoulos D. Snakes: active contour models. International
Journal of Computer Vision 1987;1:321–31.
[31] Klemencic A, Kovacic S, Pernus F. Automated segmentation of muscle fiber
images using active contour models. Cytometry 1998;32:317–26.
[32] Lam L, Suen CY. Application of majority voting to pattern recognition: an
analysis of its behavior and performance. IEEE Transactions on Systems Man
and Cybernetics, Part A: Systems and Humans 1997;27:553–68.
[33] Ko B, Seo M, Nam JY. Microscopic cell nuclei segmentation based on adaptive
attention window. Journal of Digital Imaging 2009;22:259–74.