08-09-2014, 11:28 AM
Particle Swarm Optimization Based Contrast Limited Enhancement for Mammogram Images Project Report
Particle Swarm Optimization.pdf (Size: 276.18 KB / Downloads: 22)
Abstract
In the present medical scenario detection of breast
cancer in its early stage is a very immense challenge. Many
histogram based enhancement are present today. In this
paper a Particle Swarm Optimization (PSO) for tuning the
enhancement parameter of Contrast Limited Adaptive
Histogram Equalization (CLAHE) based on Local Contrast
Modification (LCM) is presented. The PSO method of
parameter tuning adopted for LCM-CLAHE enhancement for
mammogram images achieves very good quality of images
compared to other exiting methods. The quality of enhanced
image is tested using an efficient objective criteria based on
entropy and edge information of the image. Results are
compared with other enhancement techniques such as
histogram equalization, unsharpmasking. The performance of
this method is tested using Peak Signal to Noise Ratio. The
quality of image shows that image obtained after this method
can be useful for efficient detection of breast cancer in
further process like segmentation, classification etc
INTRODUCTION
Breast cancer ranks second to lung cancer and is the most
common form of malignancy in women. Breast cancer impacts
over 240,000 new patients a year in the United States alone.
One in eight women in United States will develop breast
cancer during her lifetime [1]. 70% of breast cancer cases
occur in women who have no identifiable risk factors [2]. It
is a pernicious disease that is causing large numbers of deaths
not only in developed countries like the United States of
America, United Kingdom, Australia and Canada but also in
the underdeveloped and developing countries including
India. It occurs most commonly amongst women in the age
group of 40 to 50 years of age. As the incidence of this
disease is increasing all over the world, it is therefore an
extremely important public health objective to be able to
detect the disease at the earliest possible stage. Even with
the advancement in medical technology it is complex to
detect cancerous cells in its premature stage. In breast cancer
detection the critical part is a method to distinguish between
normal tissues and cancerous tissues. Differentiating of this
by human eye is very hard.
RELATED WORKS
Many histogram based enhancement are present today. The
Histogram Equalization (HE) which is one of the popular
methods for contrast enhancement modifies the gray level
histogram of an image to a uniform distribution [4]. But in
many cases it produces over enhancement in output image
and loss of local information which leads to insufficient
medical details during diagnosis. To overcome these
drawbacks, many variants of HE have been proposed [5-8].
In medical imaging (such as mammogram enhancement)
local contrast are more important than global contrast. In
such type of applications Global Histogram Equalization
(GHE) is insufficient because it cannot deal with local features
of original image due to its global nature.
ENHANCEMENT METHOD
HE uniformly distributes the intensity of the image .But it
produces over enhancement in the output image which leads
to the loss of local details in the mammogram images. AHE
differs from the ordinary histogram equalization in the
respect that HE generates only one histogram whereas adaptive
method computes several histogram, corresponding to a
distinct section of the image and uses that to redistribute the
intensity values of image. However, AHE has a drawback of
noise over amplification. CLAHE is a variant of AHE which
reduces the noise amplification. Using CLAHE also we have
found that it is also not so suitable for mammogram images
of very fine details. In the proposed enhancement method we
have used a local contrast enhancement (LCM) to highlight the
fine details hidden in the mammogram image and an
enhancement parameter to control the level of enhancement
along with standard CLAHE.
PSO Algorithm
The PSO is a population based search technique modeled on
the social behavior of birds within a flock. In PSO each
solution is called a particle. The PSO is an algorithm which
optimizes the particles within the search space. The individual
particles present in the space vary their position with time.
In PSO system, particles fly around in a multidimensional
search space. During flight, each particle adjusts its position
according to its own experience, and the experience of its
neighboring particles. All particles have fitness values
which are evaluated by the objective function to be
optimized, and have velocities which direct the flying of the
particles. The particles fly through the problem space by
following the personal and global best particle.
The swarm is initialized with a group of random particles
and it then searches for optima by updating through
iterations. In every iteration, each particle is updated by
following two “optima” values. The first one is the best
solution of each particle achieved so far. This value is
known as “pbest” solution. Another one is the, best solution
tracked by any particle among the whole population. This
best value is known as “gbest’ solution. These two best values
are responsible to drive the particles to move to new better
position
EXPERIMENTAL RESULTS AND DISCUSSIONS
This section presents the experimental results of the
proposed PSO based LCM-CLAHE. In this paper, the most
popular image enhancement techniques like HE, USM and
CLAHE techniques are chosen in order to validate the
proposed technique. The Local Contrast Enhancement in the
LCM-CLAHE method preserves the local information.
Determining the optimum contrast enhancement without losing
fine details is a very big challenge in mammogram contrast
enhancement. Using the PSO Optimization technique we are
getting an optimal contrast enhancement for mammogram
images as shown in Table
CONCLUSION
In this paper we have proposed an enhancement technique
called Contrast Limited Adaptive Histogram Equalization
based on Local Contrast Modification to enhance the finer
details of mammogram images and PSO optimization
technique for tuning the enhancement parameter. The proposed
method provides optimum contrast enhancement while
preserving the local information of the input mammogram
image. In our proposed method the most important property
is that it can produce better results with proper tuning of
parameter. But in case of Standard Histogram Equalization,
Unsharpmasking and Normal CLAHE it produces only one
enhanced image for a particular input image.