Download Paper Title (use style: paper title)

Detection of Skin Cancer Using Artificial Neural Network
Nishu Rani
Maneesha Nalam
Anand Mohan
M. Tech EC
Amity University
NOIDA, Uttar Pradesh
M .Tech ECE
Amity University
NOIDA, Uttar Pradesh
Assistant Professor
NSHM Group of Institutions
Durgapur .West Bengal
Abstract
Skin cancers are the most common form of cancers found in
humans. This is the most deadly form of cancer. Most of the
skin cancers are curable at initial stages. So an early
detection of skin cancer can save the patients. With the
advancement of technology, early detection of skin cancer is
possible. One such technology is the early detection of skin
cancer using Artificial Neural Network. The diagnosing
methodology uses Image processing techniques and Artificial
Intelligence. The Dermoscopy image of skin cancer is taken
and it is subjected to various pre-processing for noise removal
and image enhancement. Later image is undergone image
segmentation using Thresholding. There are certain unique
features for skin cancer regions. Such features are extracted
using feature extraction techniques like - Wavelet Coefficient
Selection 2-D method. These features are given as the input
nodes to the neural network. For classification purpose BackPropagation Neural (BPN) Network is used. It classifies the
given data set into cancerous or non-cancerous.
Keywords : Dermoscopy, Skin cancer, segmentation,
Back propagation, Wavelet transform
I.
cancer, such as: Asymmetry, Border irregularity, Color
variation and Diameter. Those are popularly known as ABCD
parameters. Asymmetry is one half of the tumour does not
match the other half. Border Irregularity is the unevenness of
images. Colour intensity change in the lesioned region is
irregular. Malignant melanoma has a diameter greater than
6mm.
II.
AUTOMATIC SKIN CANCER DETECTION
SYSTEM
Mostly automatic skin cancer detection systems consist of
following procedures:
(i) Image pre-processing – It is done to remove the noise and
fine hair.
(ii)Post-processing – It is done to enhance the shape of image.
(iii) Segmentation - It is done to remove the healthy skin from
the image and find the region of interest.
Usually the cancer cell, remains in the image; Feature
extraction extracts the useful information from the segmented
images. Finally, this information is used in the classification
system for training and testing purpose. The classification
system is usually supported by intelligent classifier, such as
neural network and support vector machine.
INTRODUCTION
Skin cancer is the uncontrolled growth of abnormal cells in the
skin. It is the cancer affecting the skin. Skin cancer may
appear as malignant or benign form. Benign Melanoma is the
least common type of skin cancer but most serious. It arises
from cancerous growth in pigmented skin lesion. Malignant
melanoma is named after the cell from which it presumably
arises, the melanocyte. If diagnosed at the right time, this
disease is curable. Its diagnosis is difficult and needs sampling
and laboratory testing. Melanoma can spread out to all parts of
the body through lymphatic system or blood. The main
problem to be considered dealing with melanoma is that, the
first affliction of the disease can pave the way for future ones.
Laboratory sampling causes the inflammation or even spread
of lesions. So, there has always been lack of less dangerous
and time-consuming methods. Computer based diagnosis can
improve the accuracy and speed of skin cancer diagnosis
which works well for disease symptoms. The similarities
among skin lesions make the diagnosis of malignant cells a
difficult task. But, there are some unique symptoms of skin
Automatic early detection system is a classification system
which distinguishes Malignant Melanoma from other skin
diseases. This methodology uses Digital Image Processing
technique and Artificial Intelligence for the classification
purpose. The input to the system is Dermoscopic images
which are in digital format. Dermoscopic images are digitally
captured and processed image. Usually such images contain
noises, so they are undergone Pre-processing to preserve the
edges. To separate the cancerous region from healthy skin,
segmentation is done. There are some unique features for the
cancerous images. Those features are extracted using Two
Dimensional Wavelet Transform in MATLAB software.
These features are given as inputs to the Artificial Neural
Network based classifier. It uses Back propagation Algorithm
for classification purpose. ANN classifies Malignant
Melanoma from Benign Melanoma. Thus detecting whether
patient is having skin cancer or not.
Dermoscopy also called as Dermatoscopy or Epiluminescence
light microscopy (ELM). It was first announced on 1987, it is
a kind of imaging technique uses to examine lesions with a
dermatoscope. The process is done by placing an oil
immersion between the skin and the optics. The lighting can
magnify the skin that improves and reveal most of the
pigmented structure, different color shades that is not visible
by naked eye and allows direct viewing and analysis of the
epidermis (the outer layer of the skin) and papillary dermis
(the deep vascular inner layer of the skin). Physician uses this
technique for diagnosis of skin cancer more efficiently. The
historical researches have proved that ELM can improve the
diagnostic accuracy by 5 – 30% compare to traditional
imaging.
Furthermore, ELM has evolution and digitalize that can be
used to classification computer. It given several advantages in
diagnosing like considering small suspicious lesion and
objective evaluation of parameters: geometry, color and
texture; and storage of image and comparable for future
development. ELM improves the accuracy of clinical
diagnosing but it requires the experienced dermatology
physician to examine the image. On Other hand, many
research for classification system has been developing over 20
years. Different approach has been used to improve the results.
Fig 1 (a) Cancerous image
Fig.1 (b) Non-Cancerous image
III.
IMAGE PROCESSING ALGORITHMS
Median filtering is a nonlinear filtering technique that is
known for preserving sharp changes in signal and for being
particularly effective in removing impulsive noise. One
effective use of median filters has been the reduction of highfrequency and impulsive noise in digital images without the
extensive blurring and edge destruction associated with linear
filters. Because the median filter is nonlinear, spectral analysis
gives little insight into the filtering process as well. Replace
each pixel value with the median value of the gray values in
the region of the pixel. It is very effective in removing salt and
pepper noise while preserving image details.
Fig.2 Computation of mean value
IV.
SEGMENTATION
Segmentation removes the healthy skin from the image and
finds the region of interest. Usually the cancer cells remains in
the image after segmentation. Segmentation used here is
Threshold Segmentation. Thresholding provides an easy and
convenient way to perform the segmentation on the basis of
the different intensities or colors in the foreground and
background regions of an image. The input to a thresholding
operation is typically a grayscale or color image. After
segmentation, the output is a binary image. Segmentation is
accomplished by scanning the whole image pixel by pixel and
labelling each pixel as object or background according to its
binarized gray level.
Segmentation algorithms are based on one of two basic
properties of intensity values discontinuity and similarity. First
category is to partition an image based on abrupt changes in
intensity, such as edges in an image. Second category is based
on partitioning an image into regions that are similar
according to predefined criteria. Histogram Threshold
approach falls under this category.
Threshold is one of the widely methods used for image
segmentation. It is useful in discriminating foreground from
the background. By selecting an adequate threshold value T,
the gray level image is converted to binary image. The binary
image should contain all of the essential information about the
position and shape of the objects of interest (foreground). The
advantage of obtaining first a binary image is that it reduces
the complexity of the data and simplifies the process of
recognition and classification. The most common way to
convert a gray-level image to a binary image is to select a
single threshold value (T). Then all the gray level values
beyond this T value will be classified as black (0), and those
above T value will be classified as white (1). The
segmentation problem becomes one of selecting the proper
value for the threshold T. A frequent method used to select T
is by analyzing the histograms of the type of images that want
to be segmented. The ideal case is when the histogram
presents only two dominant modes and a clear valley
(bimodal). In such case the value of T is selected as the valley
point between the two modes.
The histogram based techniques is dependent on the success of
the estimating the threshold value that separates the two
homogenous region of the object and background of an image.
This required that, the image formation be of two homogonous
regions and will-separated regions and there exists a threshold
value that separated these regions. The (HDT) is suitable for
image with large homogenous and will separate regions where
all area of the objects and background are homogonous and
except the area between the objects and background. This
technique can be expressed as:
C (T)=P1(T)σ12(T)+P2(T)σ22(T)
……..(I)
Where
C (T) is the within-group variance.
P1 (T) is the probability for group with values less than T.
P2 (T) is the probability for group with values greater than T.
σ12 (T) is the variance of group of pixels with values less than
Or equal to T.
σ22 (T) is the variance of group of pixels with values greater
Than T
V.
FEATURE EXTRACTION
At this stage, the important features of image data are
extracted from the segmented image. By extracting features,
the image data is narrow down to a set of features which can
distinguish between Malignant and Benign melanoma. The
extracted features should be both representatives of samples
and detailed enough to be classified. 2D wavelet transform is
used for the feature extraction. In this system, 2-D wavelet
packet is used and the enhanced image in gray scaled as an
input.
At two steps of decomposition Bior wavelet is used. At each
step of decomposition, the wavelet of primary image is
divided into an approximate and three detailed images which
show the basic information and vertical, horizontal and
diagonal details, respectively. The Features extracted using the
wavelet transform are: Mean, Standard deviation, Mean
Absolute Deviation, L1 Norm, L2 Norm.
The New Wavelet for CWT tool designs a new wavelet
adapted to a given pattern using least squares optimization.
The new wavelet can be used for accurate pattern detection
using the continuous wavelet transform (cwt). This new tool is
useful for creating new wavelets for accurate 1D pattern
detection also.
VI.
A classifier is a category of unified modeling language (UML)
elements that has some features such as attributes or methods
in common. Classifier is used for classifying Malignant
Melanoma from other skin diseases. Based on the
computational simplicity Artificial Neural Network (ANN)
based classifier is used. In this proposed system, a feed
forward multilayer network is used. Back propagation (BPN)
Algorithm is used for training. There must be input layer, at
least one hidden layer and output layer. The hidden and output
layer nodes adjust the weights value depending on the error in
classification.
In BPN the signal flow will be in feed forward direction, but
the error is back propagated and weights are updated to reduce
error. The modification of the weights is according to the
gradient of the error curve, which points in the direction to the
local minimum. Thus making it much reliable in prediction as
well as classifying tasks. BPN, weights are initialized
randomly at the beginning of training. There will be a desired
output, for which the training is done. During forward pass of
the signal, according to the initial weights and activation
function used, the network gives an output. That output is
compared with desired output. If both are not same, an error
occurs.
During reverse pass, the error is back-propagated and weights
of hidden and output layers are adjusted. The whole process
then continues until error becomes zero. The network is
trained with known values. After training, network can
perform decision making. In this methodology, Five Features
were given as input to a multilayer feed forward network.
There is one hidden layer with two hidden neurons; output
layer with one output neuron, activation function used is linear
function, which gives an output of 0 or 1. Zero represents noncancerous or benign condition and one represents cancerous or
malignant condition.
NEURAL LAB is the software used for ANN classification. It
is ANN simulation software which gives good results during
classification. The network is trained using known values of
Malignant and Benign Melanoma features. Many epochs of
training are repeated until Mean Square Error is less than
minimum value. Then data for classification is given as input
to classifier. 2 Malignant and Benign Melanoma Features were
given for classification. The output of the classifier is either
‘0’ or ‘1’. One represents cancerous condition and Zero
represents Non-cancerous condition.
ARTIFICIAL NEURAL NETWORK CLASSIFIER
Artificial neural network (ANN) architectures have been
recognized for a number of years as a powerful technology for
solving real-world image processing problems. The primary
purpose of this special issue is to demonstrate some recent
success in solving image processing problems and hopefully
to motivate other image processing researchers to utilize this
technology to solve their real-world problems.
Fig.3 Artificial Neural Network Structure.
VII. RESULTS
For smoothing image from noise, median filtering is used.
Median filtering is a common step in image processing. It is
used to minimize the influence of small structures like thin
hairs and isolated islands of pixels like small air bubbles.
Fig.7 Dimensional wavelet decomposition
FIG.4 MEDIAN FILTERED IMAGE
In this system, 2-D wavelet packet is used and the enhanced
image in gray scaled as an input. Bior wavelets at two steps of
decomposition are used. At each step of decomposition, the
wavelet of primary image is divided into an approximate and
three detailed images which show the basic information and
vertical, horizontal and diagonal details, respectively. The
Features extracted using the wavelet transform are: Mean,
Standard deviation, Mean Absolute Deviation, L1 Norm, L2
Norm.
The residual operation of the feeeded image gives the features
like: Mean, Standard deviation, Mean Absolute Deviation, L1
Norm, L2 Norm.
(a)Cancerous
Fig.8 Residual operation of Feeded image
Fig.5 Histogram of (b) Non-cancerous
Fig.6 Segmented image
Segmentation is accomplished by scanning the whole image
pixel by pixel and labelling each pixel as object or background
according to its binarized gray level.
Fig.9 Confusion matrix of class output
For the proposed system, 2Dermoscopic images were
collected from Internet. They were undergone Median
Filtering. After that, Filtered images were segmented by
Threshold Segmentation. Feature Extraction of images was
done using 2D wavelet transform. All these were done in
MATLAB software. Five features were selected for
classification- Mean, Standard Deviation, Mean Absolute
Deviation, L1 Norm, L2 Norm.
uses Digital Image Processing Techniques and Artificial
Neural Networks for the classification of Malignant
Melanoma from the other skin diseases. Dermoscopic images
were collected and they are processed using various Image
processing techniques. This methodology has got good
accuracy also. By varying the Image processing techniques
and Classifiers, the accuracy can be improved for this system
Acknowledgement
The Obtained Features were given as inputs to a Feed Forward
Neural Network. Activation function used is linear function,
which gives an output of '0' or '1'. Zero represents noncancerous or benign condition and one represents cancerous or
malignant condition. The neural network is designed using
NEURAL LAB software. The training is done with known
value. After training, Data Sets for classification were given to
the Network. 2 cases were given for classification. The
network classifies the given data into cancerous or
noncancerous. Among the 2 cases 1 was classified as
cancerous and the other as non-cancerous. It is shown in the
form of a Confusion Matrix as shown in Figure 9. It has a
good rate of accuracy too. The proposed system is proved to
be much convenient than the conventional Biopsy method.
Since this method is Computer Based Diagnosis, there is no
need for any sort of skin removal for diagnosis. It requires
only the Dermoscopic image.
Table I
N
o.
Mean
Standard
Deviation
Mean
absolute
Deviation
L2
norm
L1
norm
Output
Diagnosis
1.
230
75
45
Cancerous
203
99
80
1.55
E+05
1.46
E+05
1
2.
9.36
E+07
8.34
E+07
0
Non
Cancerous
Comparison of features of cancerous and non-cancerous image features
VIII. CONCLUSION
A Computer based early skin cancer detection system is
proposed in this paper. It proves to be a better diagnosis
method than the conventional Bioscopy method. The
diagnosing methodology uses Digital Image Processing
Techniques and Artificial Neural Networks for the
classification of Malignant Melanoma from other skin
diseases. Dermoscopic images were collected and they are
processed by various Image processing techniques. The
cancerous region is separated from healthy skin by the method
of segmentation. The unique features of the segmented images
were extracted using 2-D Wavelet Transform. Based on the
features, the images were classified as Cancerous and Noncancerous. This methodology has got good accuracy also. By
varying the Image processing techniques and Classifiers, the
accuracy can be improved for this system.
IX.
FUTURE SCOPE
A Computer based early skin cancer detection system is
proposed. It proves to be a better diagnosis method than the
conventional Bioscopy method. The diagnosing methodology
We would like to thank all faculties and students of
Electronics Department, Amity University Noida Uttar
Pradesh for their guidance and support and also facilities
provided to us.
About The Authors
Nishu Rani received the Bachelor Degree in Electronics and
communication from Shankara Institute of Technology Jaipur,
Rajasthan in the year 2011.Currently she is pursuing her
M.Tech from Amity University, Noida Uttar Pradesh.
Maneesha received the Bachelor Degree in Electronics and
communication from Amity University Noida in the year
2011.Currently she is pursuing her M.Tech from Amity
University, Noida Uttar Pradesh.
Anand Mohan
He received the MCA, M.Phil & PGDITM and also submitted
Ph.D thesis. He is currently working as an Assistant Professor
in MBA Department of NSHM Group of Institutions,
Durgapur West Bengal. His current research interest includes
VLSI & Embedded systems, Image processing, Artificial
Intelligent, Digital Instrumentation.
X.
1.
2.
3.
4.
REFERENCE
Ho Tak Lau and Adel Al-Jumaily, “Automatically
Early Detection of Skin Cancer”: Study Based on
Neural
Network
Classification,
International
Conference of Soft Computing and Pattern
Recognition, IEEE , pp 375-380, 2012
Dr. J. Abdul Jaleel, Sibi Salim, Aswin.R.B.
“International Journal of Advanced Research in
Electrical,
Electronics
and
Instrumentation
Engineering “ Vol. 1, Issue 3, September 2013
Pankaj Agrawal, S.K.Shriwastava and S.S.Limaye,
MATLAB” Implementation of Image Segmentation
Algorithms”, IEEE Pacic Rim Conference on
Communication, Computer and Signal Processing,
pp. 602-605., 2010
Ali AI-Had, M.H.Kabir Wavelets Pre-Processing of
“Artificial Neural Networks Classifiers” , IEEE
Transactions on Consumer Electronics, Vol. 53 Issue
2, pp. 593-600 , 2012.
Asadollah Shahbahrami, Jun Tang “A Colour Image
Segmentation algorithm Based on Region Growing”,
IEEE Trans. on Consumer Electronics Euromicro, pp
362-368, 2010.
6. T. Tanaka, R. Yamada, M. Tanaka, K. Shimizu, M.
Tanaka, A Study on the Image Diagnosis of
Melanoma”, IEEE Trans. on Image Processing, pp.
1010-1024, June 2004.
7. Y. Andreopoulos, N. D. Zervas, G. Lafruit, P.
Schelkens,
“A
local
wavelet
transform
implementation versus an optimal row-column
algorithm for the 2D multilevel decomposition”,
IEEE International Conf. on Image Processing,
volume 3, 2001
8. David Federal de Ouro Preto Contrast Enhancement
in “Digital Imaging using Histogram Equalization”.
IEEE International Conf. on Image Processing,
volume 54, 2012
9. Argenziano G, Soyer HP, De Giorgi V, et al.
Interactive atlas of dermoscopy. Milan: EDRA
Medical Publishing & New Media 2002.
10. Otsu N. A threshold selection method from graylevel histogram. IEEE Trans Syst Man Cybern 1979;
9(1): 62-6.
5.