A survey on enhancing mammogram image saradha arumugam academia

05/06/2015 A Survey on Enhancing Mammogram Image | saradha Arumugam Academia.edu
http://www.academia.edu/7006915/A_Survey_on_Enhancing_Mammogram_Image 1/7
A Survey on Enhancing Mammogram Image
Search... Log In Sign Up

Recent Science: International Journal of Cancer Research, ISSN:2051-784X, Vol.46, Issue.2 466
A Survey on Enhancing Mammogram Image
T.A.Sangeetha,M.Sc.,M.Phil., (PhD).,
Assistant Professor in CS,
Kongu Arts and Science College, Erode
(Research Scholar-Mother Teresa Women’s University),
Kodaikkanal, TN,India
E-Mail:tasangeetha1979@rediffmail.com.
Dr.A.Saradha
Associate Professor & H.O.D of CSE
Institute of Road and Technology,
Erode, TN, India
E-Mail: saradha@irttech.ac.in
ABSTRACT
One of the life-threatening diseases is breast cancer and it
refers to the malignancies that grow in breasts. It is the
most common type of cancer among middle aged women.
Even though, there is no guaranteed approach to prevent
breast cancer, premature detection is the key to develop
breast cancer treatment. Mammography is one of the most
successful diagnostic techniques for premature breast
cancer detection. It is a primary imaging technique for
identification and treatment of breast cancer. On the other
hand, in general the contrast of a mammogram image is
very low and it is affected with noises, in particular for
dense and glandular tissues. In these cases the radiologist
possibly will miss certain diagnostically significant
microcalcifications. With the purpose of improving
diagnosis of cancer accurately at the premature stage,
image enhancement technology is often used which assist
in enhancing the image. Being the fact that these
mammograms are low contrast, blur and fuzzy, it is
extremely complicated to identify the microcalcifications.
Mammogram enhancement is necessary for the
enhancement of contrast features and to remove noise.
Proper image enhancement technique improves the
visibility of microcalcifications. The fundamental
requirement in mammogram enhancement is to enhance
its contrast. Several numbers of studies have been carried
out with the intention of enhancing the mammogram
images which are discussed in the literature.
Keywords---Digital Mammogram, Breast Cancer, Image
Enhancement, Microcalcifications, Contrast
Enhancement.
1. INTRODUCTION
A malignant tumor developed from breast cells is called
breast cancer. It is one of the most fatal diseases for
middle-aged women and it is one of the leading causes of
women mortality. One among eight women is affected
with this disease[1]. Premature detection is the best
possible way to improve breast cancer diagnosis because
the causes of the disease are still indefinite. Premature
detection is the key solution for improving breast cancer
diagnosis.
tumors like microcalcifications, masses and stellate
lesions [2]. Based on the medical perspective, the most
primitive symptom of breast cancer is the emergence of
microcalcifications. As a result, the detection of
microcalcification is a major part of diagnosis in early
stage breast cancer. On the other hand, microcalcification
is extremely small to identify. Mammography is known as
the best modality to identify microcalcification [3].
The tiny size of microcalcification results in unclear
visualization in mammograms. As a result, to provide the
enhanced visibility of breast cancer to physicians in
addition to automatic breast-cancer detection systems,
mammogram contrast should be enhanced [4]. In
mammograms, the size of microcalcification is almost
nearer to noise. In order to identify the breast cancer at
premature stage, it is necessary to reduce the noise level at
the same time enhance the microcalcification area.
In recent past, mammographic interpretation was
supported by computer-based techniques which are
exploited either as visualization tools or as estimation
devices [5].
Imaging techniques play a significant role in helping
perform digital mammogram, especially of abnormal
areas that cannot be felt but can be seen on a conventionalmammogram. Before any image-processing algorithm of
mammogram pre-processing steps are very important in
order to limit the search for abnormalities without undue
influence from background of the mammogram [6].
Image enhancement algorithms have been exploited for
the enhancement of contrast features and the removal of
noise. At first, contrast enhancement approaches have
exploited the convolution function or techniques such as
morphological, edge detection and band-pass filters [7, 8].
The enhancement approaches could be classified based on
the generalization of their parameters as global, local or
adaptive [9, 10].
The most significant objective of mammography is to
identify small, non-palpable cancers in its premature
saradha Arumugam  213  Info Download Uploaded by
PDF


©Recent Science Publications Archives 2012|$25.00 | 10004211 |
diagnosis.
The success of diagnosis is completely based on the early
detection of the disease. At present Breast cancer
identification is carried out on mammograms by
radiologists by using a magnifying glass to observe
identify small, non-palpable cancers in its prematurestage. However, mammograms are complicated to
understand as the pathological transformations of the
breast are subtle and their visibility is poor in low contrast
and noisy mammograms. With the purpose of increasing
the visibility of features several approaches for image
enhancement have been proposed which are discussed in

the literature. Contrast enhancement is a necessary
technical assistance in applications where human visual
perception remains the most important approach to obtain
appropriate information from images.
The significant features of the mammogram images have
to be enhanced to obtain the information regarding the
unknown characteristics. Mainly, image enhancement
includes intensity and contrast manipulation, noise
reduction, background removal, edges sharpening,
filtering, etc. Mammogram enhancement approaches are
employed for increasing identification, characterization
efficiency and also as preprocessing stage. Several
processing techniques for contrast enhancing in the region
of microcalcifications have been reported in the following
section.
2. LITERATURE SURVEY
In breast cancer diagnosis, the radiologists mostly use
their eyes to separate cancer when they monitor the
mammograms. Still, in many cases, cancer is not simply
detected by the eyes because of the bad imaging situation.
In order to expand the exact study rate of cancer, image-
enhancement knowledge is often used to recover the
image and assist the radiologists. Jinshan Tang et al., [11]
developed a new image-enhancement technology in the
wavelet domain for radiologists to screen mammograms.
The original image-enhancement algorithm has some
compensation. First, the projected image-enhancement
skill modifies a multiscale assess which matches the
human vision method and thus the improved images have
better visual quality; second, the image improvement is
competent in the wavelet area and therefore it preserve
time keeping if the image is solid by wavelet revise based
methods; third, the end users can accurate the
improvement by manipulating a single control.
Panetta et al., [12] introduced a new unsharp masking
(UM) scheme, called nonlinear UM (NLUM), for
mammogram enhancement. The NLUM offers the
suppleness 1) to entrench particular types of filters into
the nonlinear filtering operative; 2) to decide various
linear or nonlinear operations for the combination of
processes that joins the better filtered sector of the
mammogram with the original mammogram; and 3) to let
the NLUM parameter collection to be performed
physically or by using a quantitative development assess
to get the best enhancement parameters. The author
created an improvement assess advance, called the
second-derivative-like assess of improvement, which is
revealed to contain better presentation than additional
events in evaluating the visual quality of image
improvement. The relationship and valuation of
improvement.
Tang et al., [13] proposed a novel algorithm for
multidimensional image enhancement based on a fuzzy
domain enhancement method, and an implementation of a
recursive and separable low-pass filter. Allowing for a
curved image as a fuzzy data set, every pixel in an image
is processed separately, with fuzzy area alteration and
enhancement of both the lively sort and the limited gray
altitude variations. The algorithm has the reward of being
rapid and adaptive, so it can be used in real-time image
dispensation applications and for multidimensional data
with small computational cost. It also has the capability to
decrease noise and unnecessary conditions that may
involve the visualization superiority of two-dimensional
(2-D)/three-dimensional (3-D) data. Some of the examples
of the algorithm are given for mammograms, ultrasound
3-D images, and clear images.
Analytical features in mammograms differ extensively in
size and shape. Traditional image improvement
techniques cannot adjust to the unstable characteristics of
analytical features. An adaptive technique for enhancing
the contrast of mammographic features of varying size
and shape was presented by Morrow et al., [14]. The
technique uses every pixel in the image as a seed to
develop an area. The level and outline of the region adjust
to local image gray-level variations, matching to an image
feature. The variation of each region is designed with
respect to its entity conditions. Distinction is then
improved by applying a practical alteration based on
every region's germ pixel price, its compare, and its
conditions. A quantity of assessment of image compare
improvement is also clear based on a histogram of region
contrast and used for evaluation of results. Using
mammogram images digitized at high decision (less than
0.1 mm pixel size), it is revealed that the strength of
microcalcification clusters and anatomic details is greatly
improved in the processed images.
Singh et al., [15] proposed a novel set of metrics that
measure the quality of the image enhancement of
mammographic images in a computer-aided detection
framework aimed at automatically finding masses using
machine learning techniques. The methodology includes a
new method for the grouping of the metrics projected into
a single quantitative measure. The author evaluated
methodology on more than 150 images from the widely
offered digital database for monitoring mammograms.
The author showed that the quantitative actions help to
choose the finest suitable image improvement on a per
mammogram basis, which improves the value of
subsequent image segmentation much superior than using
the same improvement method for all mammograms.

improvement. The relationship and valuation of
improvement presentation reveal that the NLUM can
develop the disease diagnosis by enhancing the well
details in mammograms with no priori information of the
image contents. The human image method based image
rot is used for learning and image of mammogram
the same improvement method for all mammograms.
Wei Qian et al., [16] described an article about the
application of wavelet transform for image enhancement
in medical imaging. The first scientific application is the
improvement of microcalcification clusters (MCCs) in

digitized mammograms to get better in both the
visualization and the finding using computer assisted
diagnostic (CAD) methods. The possible widespread
function for improved visual interpretation of medical
images using a computer monitor is also demonstrated.
The early detection of MCCs is important in screening
programs since their presence is often associated with a
high incidence of breast cancer. The improvement of
MCCs is an outstanding model for real world estimation
of the wavelet transform. The finding of MCCs presents
an important confront to the presentation characteristics of
X-ray imaging sensors and image present monitors ever
since microcalcifications vary in size, shape, signal
intensity, and contrast and may be situated in areas of very
solid parenchymal tissue, building the discovery difficult.
The categorization of MCCs, in turn, as kind or evil,
requires the morphology and detail to be sealed.
Heinlein et al., [17] presented a new algorithm for
enhancement of microcalcifications in mammograms. The
major innovation is the function of methods that have
urbanized for creation of filterbanks resulting from the
regular wavelet transform. These separate wavelet
decompositions, called incorporated wavelets, are
optimally intended for improvement of multiscale
structures in images. Moreover, a model based approach
is used to clean presented methods for common
improvement of mammograms resulting in a more exact
improvement of microcalcifications. The author proposed
results of the method and contrast them with standard
algorithms. Finally, it indicates how these techniques can
also be useful to the finding of microcalcifications.
Peripheral enhancement and tilt correction of unprocessed
digital mammograms was achieved by Snoeren et al., [18]
with a new reversible algorithm. This technique has two
major rewards for image visualization. (1) the displayed
active range can be comparatively small, and (2) alteration
of the overall luminance to examine information is not
necessary in most cases. The rectification is helpful for
preprocessing in computer-aided discovery/analysis
algorithms. The technique is based on information of the
three-dimensional compacted breast shape to make equal
width by adding virtual tissue, which results in force
equalization for the mammographic image. Earlier
described methods completely guess the charitable of
width variations to image strength, frequently by
nonparametric methods. The projected technique provides
a worldwide parametric breast figure model, which is
beneficial for image and CAD.
Mammograms are difficult to interpret, especially of
cancers at their early stages. Rangayyan et al., [19]
digitized with a high-resolution of about 4096×2048×10-
bit pixels and then processed with the ANCE method.
Unrefined and refined digitized mammograms as well as
the new films were obtainable to six knowledgeable
radiologists for a receiver operating characteristic (ROC)
assessment for the hard case set and to three mention
radiologists for the period cancer set. The results explain
that the radiologists' presentation with the ANCE-
processed images is the greatest amongst the three sets of
images (original, digitized, and enhanced) in conditions of
area under the ROC curve and that analytic sympathy is
enhanced by the ANCE algorithm. All of the 19 period
cancer cases not detected with the new films of previous
mammographic examinations were diagnosed as
malignant with the consequent ANCE-processed versions,
while only one of the six kind cases at first labeled
properly with the original mammograms was interpreted
as malignant after improvement. This learning
demonstrates the possible for development of diagnostic
presentation in early discovery of breast cancer with
digital image enhancement.
Unser et al., [20] presented an overview of the various
uses of the Wavelet Transform (WT) in medicine and
biology. The author described the wavelet kinds that are
mostly significant for biomedical application. Above all it
provides an accepting of the Continuous Wavelet
Transform (CWT) as a prewhitening multiscale
synchronized filter. It also in short specifies the
comparison among the WT and few of the organic
processing that occurs in the early mechanism of the aural
and visual system. It then recorrects the uses of the WT
for the examination of 1-D physiological signals obtained
by phonocardiography, electrocardiography (ECG), mid
electroencephalography (EEG), as well as evoked reaction
potentials. Next, it provides a review of wavelet
developments in remedial imaging. These contain
biomedical image giving out algorithms (e.g., noise
reduction, image enhancement, and detection of
microcalcifications in mammograms), image rebuilding
and gaining schemes (tomography, and Magnetic
Resonance Imaging (MRI)), and multiresolution methods
for the register and statistical study of efficient images of
the brain (Positron Emission Tomography (PET) and
functional MRI (fMRI)). In each case, it provides the
reader with same common background information and a
brief clarification of how the methods work.
Mammography is the well-organized method for the early
finding of breast diseases. On the other hand, the classic
analytic signs like microcalcifications and loads are hard
to distinguish because mammograms are low-contrast and

analyzed the effectiveness of Adaptive Neighborhood
Contrast Enhancement (ANCE) technique in increasing
the sensitivity of breast cancer diagnosis. Seventy-eight
screen-film mammograms of 21 hard cases (14 benign
and seven malignant), 222 screen-film mammograms of
28 period cancer patients and six kind organize cases were
noisy images. Mencattini et al., [21] proposed a novel
algorithm for image denoising and enhancement based on
dyadic wavelet processing. The denoising stage is based
on the limited iterative noise difference evaluation. In
addition, in the case of microcalcifications, the author
proposed an adaptive change of improvement degree at

A survey on enhancing mammogram image saradha arumugam academia

More Related Content

What's hot (20)

Viewers also liked (14)

Similar to A survey on enhancing mammogram image saradha arumugam academia (20)

Recently uploaded (20)

A survey on enhancing mammogram image saradha arumugam academia