Spatial Domain Enhancement Techniques For Detection of Lung Tumor From Chest X-Ray Image

International Journal of Application or Innovation in Engineering& Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com

Volume 2, Issue 8, August 2013 ISSN 2319 - 4847

Volume 2, Issue 8, August 2013 Page 276

Abstract
Advances in the area of bio-medical image processing of chest X-ray has resulted in the acquisition of high-quality images of the
human chest. With these advances, there arises a genuine need for image processing algorithms specific to the chest, in order to
fully exploit this digital technology. Image enhancement is an important part of image processing and it is applied in every field
where images are ought to be understood and analyzed. Image enhancement is basically improving the interpretability or
perception of information in images for human viewers and providing better input for other automated image processing
techniques. The principal objective of image enhancement is to modify attributes of an image to make it more suitable for a given
task and a specific observer. During this process, one or more attributes of the image are modified. In image enhancement,
resizing, improving brightness and contrast of the image, filtering is done. The Enhancement process improves the visual quality
of the image by removing extraneous information from the image. The paper focuses on spatial domain techniques for image
enhancement, with particular reference to point processing methods and histogram processing.

Keywords: Negativity, Profile, Enhancement, Radiologist
1. INTRODUCTION
The success of digital techniques in various medical imaging modalities (computed tomography, nuclear magnetic
resonance, and digital subtraction angiography) has generated considerable interest in chest radiology. One of the primary
reasons for this trend toward digital chest radiography is that the human chest is a very difficult structure to image [1].
The major challenge is the wide dynamic range of information between X-rays emerging from the heavily attenuated
mediastinum (heart, spine, aorta, and other central features on the radiograph) and those that have passed through the
air-filled lungs. Although this wide dynamic range is difficult to represent faithfully with current film-screen techniques,
many of these problems could be circumvented by a digital representation [2]-[7].
Lung cancer is known as one of the most difficult cancers to cure and the number of deaths that it causes is generally
increasing. A detection of the lung cancer in its early stage can be helpful for medical treatment to limit the danger [8].
One of the measures is a mass screening process for lung cancer. As a conventional method for mass screening process,
chest X-ray films have been used for lung cancer diagnosis.
In the lung, either malignant and benign tissues or one of them is present. It became difficult for radiologist to decide
whether the patient is having malignant tissues or not. Malignant tissues are cancerous and benign tissues are not
cancerous. Not only the tissues are creating problems in analyzing chest X-ray, but also the presence of noise, overlapping
of the organ and the presence of any other disease (cough, T.B. etc.). Sometimes visual and technical limitation makes it
difficult to detect lung cancer.
X-ray techniques are economical and required less time to generate X-ray image which makes it popular among the
middle class families and radiologist. But on the other hand it has some limitation. Discovering a lung nodule from a
chest X-ray image can be challenging task and missed due to a variety of factors including the presence of ribs, clavicle,
trachea, blood vessels, and heart which could block or obscure the view of a nodule [9]. Also, suboptimal image
acquisition and ill-defined margins compound the difficulty for radiological interpretations. Digital image processing
offers various enhancement techniques which could facilitate a more accurate detection of a lung nodule on a chest
radiograph. Image enhancement is used to improve the interpretability or perception of information in images for human
viewers, or to provide better input for other automated image processing techniques [10].
In this study, the paper focus on spatial domain point processing-filtering, histogram and contrast enhancement
techniques which are to be used to improve the visibility of a chest radiograph and to make a lung nodule more evidently
seen through reducing relative tonal background, sharper edges, and enhanced contrast. The study serves to facilitate
early detection of a lung nodule from a chest radiograph as a pre-screening procedure.
Spatial Domain Enhancement Techniques for
Detection of Lung Tumor from
Chest X-Ray Image

Manoj R. Tarambale
1
and Nitin S. Lingayat
2

1
Electrical HOD, Marathwada Mitra Mandals college of Engineering,
Pune, Pin-411052, University of Pune, Maharashtra, India

2
Electrical HOD, DR. Babasaheb Ambedkar Technological Universitys Institute of Petrochemical Engineering,
Lonere, Maharashtra, India Pin402103


2. FILTER
Filtering in image processing is a process that cleans up appearances and allows for selective highlighting of specific
information. Noise is basically considered as a high pass region. For removing noise low pass filter is used. No matter
how much care one takes, some amount of noise always creeps in the image [11]. Noises which are commonly found in
any image are Gaussian noise and salt & pepper noise.

2.1 Averaging filter
It is a spatial domain filter used for noise reduction. The output of the average filter is simply the average of pixel
contained in the neighbor of the filter mask. The size of the neighborhood controls the amount of filtering. In averaging
filter, each pixel is replaced by a weighted average of its neighborhood pixels. It preserves the smooth region in the image
and its removes the sharp variation leading to blurring effect. Salt and pepper noise are a type of the noise that are
commonly found in the image. Averaging filter eliminates salt and pepper noise. Average filtering is given by equation:-
) 1 (
) , (
) , (
1
) , (
=
xy
S t s
t s g
n m
y x f

Mask for averaging filter is given by

Figure 1 33 averaging filter mask

If the average of the mean of the whole image and mean of the neighboring pixel is computed then the image sharpness
can be retained. It is given by

) 2 (
2
) , (
1
) , (
1
) , (
) , ( ) , (

e e

+
=
xy xy S t s S t s
t s g
n m
t s i
N M
y x I

Where ) , (
y x I is a resulted image. M and N is the number of rows and columns of the image. m and n are the
number of rows and columns of the mask. ) , ( t s i is the original image and ) , ( t s g is the image obtain after mask on the
image. It will reduce the blurring of the image.

2.2 High Boost filter
High boost filter is a modified version of the high pass filter. It is often desirable to emphasize high frequency
components representing the image details without eliminating low frequency components (such as sharpening). In this
case, the high-boost filter can be used to enhance high frequency component while still keeping the low frequency
components:

High pass image =original image low pass image (3)
original highboost
original highpass allpass
highpass original highboost
I W
I W cW
I cI I
*
* ) ( ) 4 (
=
+ =
+ =

) 5 (
highpass allpass highboost
W cW W + =
Where c is a constant.
highboost
W mask of the high boost filter.
=
highboost
W
(
(
(
+

=
(
(
(

+
(
(
(
0 1 0
1 4 1
0 1 0
0 1 0
1 4 1
0 1 0
0 0 0
0 1 0
0 0 0
c c

=
highboost
W
(
(
(

+

=
(
(
(

+
(
(
(
1 1 1
1 8 1
1 1 1
1 1 1
1 8 1
1 1 1
0 0 0
0 1 0
0 0 0
c c

If c >0 then some of the original signal is added back to the high pass result. This process restores background in the
high passed image.


3 ENHANCEMENT
3.1 Negativity
Negative simply means inverting the gray level i.e. if pixel in the original image is black, then it will look white and vice
versa. It shifts histogram from the brighter side to darker side or darker side to the brighter side. In general

s =(L - 1) r (6)

Where L is a maximum gray value in the image. r is the gray value of the pixel.

Figure 2 Negative transform for an 8 bit image

In Figure 2, vertical axis represent modified grey level s and horizontal axis represent an original grey level r.

3.2 Contrast Manipulation
Low contrast images occur due to poor or non-uniform lighting conditions or due to nonlinearity or small dynamic range
of the imaging sensor. Horizontal axis r represents the input pixel value, and the vertical axiss represents the output
pixel value. As seen, there are three straight line segments used to transform an input pixel to its resulting output pixel
value. The transformation of the input pixel value to the output pixel value is via the piecewise linear profile shown in the
Figure 3.
The parameters specifying the contrast stretch mapping are the four values r2, s2, r3, s3, which determine the position of
the intermediate straight line segment. Modifying any of these four values modifies the contrast stretch transformation.
The values of r1, s1, r4, s4 are fixed.
a and b are the two thresholds. l, m and n are the slopes of the straight line in contrast streching transform.

Figure 3 Contrast Stretching Transformation. For dark region stretch l >1, mid region stretch m >1, bright region
stretch n >1.
Formula for contrast stretching algorithm is given by
) 7 (
1
0
) .(
) .(
.
< s
s s
< s
+
+ =
L r b
b r a
a r
w b r n
v a r m
r l
s

The parameter a and b can be obtained by examining the histogram of the image.

3.3 Histogram and Histogram Manipulation
The histogram provides a convenient summary of the intensities in an image. The histogram of an image is a plot of the
number of occurrences of gray levels in the image against the gray-level values. The horizontal axis of the plot represents
the tonal variations, while the vertical axis represents the number of pixels in that particular tone. The left side of the
horizontal axis represents the black and dark areas, the middle represents the medium grey and the right hand side
represents light and pure white areas. The vertical axis represents the size of the area that is captured in each one of these
zones [12].
The luminance histogram of a typical natural scene that has been linearly quantized is usually highly skewed toward the
darker levels; a majority of the pixels possesses a luminance less than the average. In such images, detail in the darker


regions is often not perceptible. Histogram manipulation basically modified the histogram of an input image as to
improve the visual quality of the image. It is process that attempt to spread out the gray levels in an image so that they are
evenly distributed across their range. It reassigns the brightness values of pixels based on the image histogram and
provides more visually pleasing results across a wider range of images.

4 SEGMENTATION
Segmentation partitions an image into distinct regions containing each pixel with similar attributes. To be meaningful
and useful for image analysis and interpretation, the regions should strongly relate to depict objects or features of interest.
Meaningful segmentation is the first step from low-level image processing transforming a grayscale or colour image into
one or more other images to high-level image description in terms of features, objects, and scenes. The success of image
analysis depends on the reliability of segmentation, but an accurate partitioning of an image is generally a very
challenging problem.

4.1 THRESHOLDING
It is the simplest non-contextual segmentation technique. With a single threshold, it transforms a grayscale or colour
image into a binary image considered as a binary region map. The binary map contains two possibly disjoint regions, one
of them containing pixels with input data values smaller than a threshold and another relating to the input values that are
at or above the threshold. The former and latter regions are usually labeled with zero (0) and non-zero (1) labels,
respectively. The segmentation depends on image property being thresholded and on how the threshold is chosen [12].
Generally, the non-contextual thresholding may involve two or more thresholds as well as produce more than two types of
regions such that ranges of input image signals related to each region type are separated with thresholds. The question of
thresholding is how to automatically determine the threshold value.

The most common image property to threshold is pixel grey level:
) 8 (
) , ( 0
) , ( 1
) , (
(
s
>
=
T y x f if
T y x f if
y x g

where T is the threshold.

Using two thresholds, T1 <T1, a range of grey levels related to region 1 can be defined:
) 9 (
1
2 1
2
) , (
) , (
) , (
) , (
(
(
(
s
s <
>
=
T y x f if c
T y x f T if b
T y x f if a
y x g

Where a, b and c are any three distinct intensity values and T1, T2 are the two thresholds [14], [15].

4.2 Morphological Image Processing
It is a collection of non-linear operations related to the shape or morphology of features in an image. Morphological
operations rely only on the relative ordering of pixel values, not on their numerical values, and therefore are especially
suited to the processing of binary images. Morphological operations can also be applied to grayscale images such that
their light transfer functions are unknown and therefore their absolute pixel values are of no or minor interest.
Morphological techniques probe an image with a small shape or template called a structuring element. The structuring
element is positioned at all possible locations in the image and it is compared with the corresponding neighborhood of
pixels. Some operations test whether the element "fits" within the neighborhood, while others test whether it "hits" or
intersects the neighborhood.
A morphological operation on a binary image creates a new binary image in which the pixel has a non-zero value only if
the test is successful at that location in the input image.

4.3 Dilation
It is a process in which the binary image is expanded from its original shape. The way binary image is expanded is
determined by the structuring element. The structuring element is smaller in size compared to the image itself, and
normally the size used for the structuring element is 33.the structuring element is reflected and shifted from left to right
and from top to bottom, at each shift the process will look for any overlapping similar pixels between the structuring
element and that of binary image. If there exists an overlapping then the pixels under the central position of the
structuring element will be turned to 1 or black. It has effects of filling in the valleys between spiky edges [13].

4.4 Erosion
It is the process in which the binary image shrinks and the shrunk is determined by the structuring element. The
structuring element is smaller in size compared to the image itself, and normally the size used for the structuring element
is 33.the structuring element is reflected and shifted from left to right and from top to bottom, at each shift the process


will look for, whether there is a complete overlap with the structuring element or not. If there is no complete overlapping
then the centre pixel indicated by the centre of the structuring element will be set white or 0. It has the effect of deleting
spiky edges [13].

5 BLOCK DIAGRAM

Figure 4 Stages of propose algorithm

Figure 2 Indicates various steps used in digital image processing. These steps are followed for enhancing and
segmentation of the chest X-ray image.

6 RESULT AND DISCUSSION
The success of digital techniques in various medical imaging modalities (X-rays, CTs, MRI and PET scan) has generated
considerable interest in chest radiology. But their success in detecting lung cancer in its early stage is limited to certain
levels. A Lot of research is carried out in the field of image processing for early detection of cancer. Because of the
environment harsh condition, human and technical error affects the quality of the image and produces a difficulty in for
radiological interpretation. Digital image processing facilities various algorithms, which help in improving image quality
and makes image error free. It improves the perception of information in image for human viewers. In digital image
processing, image parameters are manipulated, in order to obtain certain useful information for further classification.
Detection of the tumor is difficult because of-
1. Presence of noise (salt and pepper, Gaussian
noise)
2. Overlapping of body organs (rib cage)
3. Presence of benign tissues.
4. Patient suffering from lung diseases (tuberculosis or cough)
5. Limitation of human visualization

In figure 5 shows the original chest X-ray image which is obtained after scanning. The Image is too bright to visualize
and difficult to localized lung tumor.

Figure 5 Original image and its histogram

For better visualization, image negativity is found. The histogram shows that pixels are shifted from the brighter region
(right hand side) to the darker region (left hand side) as shown in figure 6.


Figure 6 Negative image and its histogram

Figure 7 Noisy image and its histogram

Noisy image and its histogram are shown in figure 7. Salt and pepper noise contain a pixel value either 0 or 255. In the
histogram long lines can be seen on 0 and 255 gray level. Filtered image and histogram in figure 8, shows that the
number of pixels at 0 and 255 gray level value is reduced.

Figure 8 Filtered Image and its histogram

Filtering causes blurring and shifts the histogram slightly left. For overcoming the problem mean of the original image is
calculated. Then the average of the pixel value of filtered image and the means is calculated. Calculation result is shown
in figure 9 .

Figure 9 Image obtain after averaging image gray level with image mean. Histogram of the image



Figure 10 Contrast Manipulated image and its histogram

Figure 11 Histogram equalization image and its histogram

After doing averaging still some information is hidden. Contrast and histogram manipulation is used to improve the
contrast and the luminance of the image. Now the image is enhanced and the segmentation techniques can be applied for
the segmenting interested area. Figure 12 shows the threshold image. For segmenting image two thresholds are used.

Figure 12 Thresholding image

Figure 13 Eroded image


Figure 14 Dilated image

Figure 13 and 14 shows eroded and dilated image. These are the basic morphological operation. Thresholding shows
interested area is attached with the rest of the unwanted structure. Erosion and dilation made it possible. Erosion reduces
the size of the structure and dilation increase the size of the reduce structure. By doing this interested structure segment
out from the rest of the structure.

Figure 15 Label Image

For separating the interested area from the rest of the unwanted structure label algorithm is used. In this algorithm each
separated structure is represented by a number. By selecting this number interested area can be separate out. Figure 15
shown binary image and its structure are labeled with numbers. Figure 16 shows separated structure.

Figure 16 Segmented Suspicious Regions

Figure 7 is obtained by superimposing figure 16 image on the original image. By doing this area is segmented from the
original image


Figure 17 Segmented image Superimposed on the original image

7. ACKNOWELGMENT
We would like to express our deepest appreciation to the Japanese Society of Radiological Technology (JSRT) in
cooperation with the Japanese Radiological Society (JRS) for providing clinically well proven images for research
purpose.

8. CONCLUSION
In the digital image processing, a lot of research is carried for the precisely segmenting lung tumor from the X-ray image.
Human errors and overlapping of mediastinum structure makes the task of segmentation difficult. The algorithm
described in the paper is simple and easy to implement. These simple techniques are precisely segmenting suspicious
region from the original image. Enhancement helps in bringing forward the useful important detail from the image by
reducing irrelevant information. Further, feature extraction and classification techniques decided whether the segmented
region is suspicious or not.

REFERENCE

[1] Robert H. Sherrier and G. A. Johnson, Regionally Adaptive Histogram Equalization of the Chest IEEE
Transactions On Medical Imaging, vol. 6, no. 1, March 1987.
[2] G. A. Johnson and C. E. Ravin, "A survey of digital chest radiography," Radiol. Clin. North Amer., vol. 21, pp. 655-
665, 1983.
[3] G. A. Johnson, N. Danieley, and C. E. Ravin, "Processing alternatives for digital chest imaging," Radiol. Clin. North
Amer., vol. 23, pp. 244, 1985.
[4] R. G. Fraser, E. Breatnach, and G. T. Barnes, "Digital radiography of the chest: Clinical experience with a prototype
unit," Radiology, vol. 148, pp. 1-5, 1983.
[5] M. Sonoda, M. Takano, J. Miyahara, and H. Kato, "Computed radiography utilizing scanning laser stimulated
luminescence," Radiology, vol. 148, pp. 833-838, 1983.
[6] L. R. Goodman, W. D. Foley, C. R. Wilson, A. A. Rimm, and T.L. Lawson, "Digital and conventional chest images:
Observer performance with film digital radiography system," Radiology, vol. 158, pp. 27-33, 1986.
[7] D. P. Chakraborty, E. S. Breatnach, M. V. Yester, B. Soto, G. T.Barnes, and R. G. Fraser, "Digital and conventional
chest imaging: A modified ROC study of observer performance using simulated nodules," Radiology, vol. 158, pp.
35-39, 1986.
[8] Radhika sivaramakrishna, J. Scott Birbeck, Brent E. Higgs, and Robrt M. cothren, Medical Computer-aided
Detection of Lung cancer from Computed Tomography Images, Technology review journal* Fall/ Winter 2002.
[9] J. H. M. Austin, B. M. Romney, and L. S. Goldsmith, Missed bronchogenic carcinoma: Radiographic findings in 27
patients with a potentially resectable lesion evident in retrospect, Radiology, vol. 182, pp. 115-122, January 1992
[10] Benjamin Y. M. Kwan And Hon Keung Kwan, Improved Lung Nodule Visualization On Chest Radiographs Using
Digital Filtering And Contrast Enhancement, World Academy Of Science, Engineering And Technology, Issue- 60,
pp-590-593, Dec 2011.
[11] Fabrizio Russo, An Image Enhancement Technique Combining Sharpening and Noise Reduction, IEEE
Transactions On Instrumentation And Measurement, vol. 51, no. 4, August 2002
[12] Besma Abidi and Mark Mitckes, Grayscale enhancement techniques of x-ray images of carry-on luggage,
International conference on quality control by artificial intelligence, vo.5132, pp.579-591, May 2013.


[13] S.Jayaraman and T.Veerakumar, Digital Image Processing, 3rd ed., New Delhi, India: Tata McGraw Hill, 2010,
ch. 5, pp. 394-395.
[14] Rafael C. Gonzalez, and Richard E. Woods, Digital Image Processing,. 3rd edition, Prentice Hall, 2008, ch.10,
pp.760-763.
[15] Rafael C. Gonzalez, and Richard E. Woods, Digital Image Processing using MATLAB,. 2nd edition, Prentice Hall,
2009.
[16] A.K. Jain, Fundamental of Digital Image Processing, Prentice Hall, Englewood Cliffs, NJ.1989.

AUTHOR
Manoj R. Tarambale was born in Jalgaon District, India, in 1970. He received graduate degree (B.E.) in
Electrical Engineering from BVCOE, Pune-43, University of Pune, India, in 1992 and post graduate degree
(M.E.) in Control System from WCOE, Sangli, Shivaji University, Kolhapur, India in 2002. Currently, he is
pursuing his Ph.D. degree from PACIFIC University, Udaipur, India.
He has one year industrial experience and twenty years teaching experience. At present, he is Assistant Professor and
Head of electrical engineering department of Marathwada Mitra Mandals College of Engineering, Pune-52, University of
Pune, India. His main research interests are bio-medical image processing, electronic instrumentation, process control
instrumentation and medical diagnosis.
Prof. Tarambale is a member of Institute of Engineers (India) and Indian Society for Technical Education (ISTE).

Nitin S. Lingayat was born in Shahapur (Thane), India in 1971. He received the B.E. degree in Electrical
Engineering from the University of Poona, Pune in 1992 and the M.Tech degree from Indian Institute of
Technology Bombay, Mumbai in 1998. He obtained Ph.D. degree in Electrical Engineering from J.N. Vyas
University, Jodhpur, India in 2008. He is Head, Electrical Engineering Department, Institute of Petrochemical
Engineering of Dr. Babasaheb Ambedkar Technological University, Lonere, Maharashtra (India) from January 2004.
He has a teaching experience of twenty one years. His research interest includes biomedical signal and image processing,
energy management, statistical signal processing, non-conventional energy sources etc.
Dr. Lingayat is a member of IAENG, Institute of Engineers (India), Indian Society for Technical Education (ISTE). He
has published many papers in the National and International Journals.

Spatial Domain Enhancement Techniques For Detection of Lung Tumor From Chest X-Ray Image

Uploaded by

Copyright:

Available Formats

Spatial Domain Enhancement Techniques For Detection of Lung Tumor From Chest X-Ray Image

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Spatial Domain Enhancement Techniques For Detection of Lung Tumor From Chest X-Ray Image

Uploaded by

Copyright:

Available Formats

International Journal of Application or Innovation in Engineering& Management (IJAIEM)

Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.