Skin Cancer Detection Using Image Processing

Abstract

Skin cancer is one of the most popular types of cancer, which inspires the life of millions of people every year in the entire world. Melanoma is one of the forms of cancer that initiates in melanocytes and it can influence the skin only. It’s more serious as compare with other types of skin cancer. The Melanoma can be of benign or malignant. The paper focused on detection system has been designed for diagnosing melanoma in early stages by using digital image processing techniques. The paper has many steps like preprocessing, segmentation, feature extraction and detection process which give the acceptable results for skin cancer detection problems. In today’s modern world, Skin cancer is the most common cause of death amongst humans. Skin cancer is abnormal growth of skin cells most often develops on body exposed to the sunlight, but can occur anywhere on the body. Most of the skin cancers are curable at early stages. So an early and fast detection of skin cancer can save the patient’s life. With the new technology, early detection of skin cancer is possible at initial stage. Formal method for diagnosis skin cancer detection is Biopsy method.

Introduction

I. INTRODUCTION

Skin cancer is a malignant tumour which grows in skin cells. It is one of the most common of all cancer which affects human beings and accounts for more than 50% of all types of cancers around the world. Skin cancer is skin’s unwanted growth with differing causes and varying degrees of malignancies. It can spread very fast to all organs/parts of human body through lymphatic system or blood. The incidences of melanoma - the deadliest form of skin cancer has been on rise at an alarming rate of 3% per year.

Skin cancer is a deadly disease. Skin has three basic layers. Skin cancer begins in outermost layer, which is made up of first layer squamous cells, second layer basal cells, and innermost or third layer melanocytes cell. Squamous cell and basal cell are sometimes called non-melanoma cancers. Non-melanoma skin cancer always responds to treatment and rarely spreads to other skin tissues. Melanoma is more dangerous than most other types of skin cancer .If it is not detected at beginning stage, it is quickly invade nearby tissues and spread to other parts of the body. Formal diagnosis method to skin cancer detection is Biopsy method. A biopsy is a method to remove a piece of tissue or a sample of cells from patient body so that it can be analysed in a laboratory. It is uncomfortable method. Biopsy Method is time consuming for patient as well as doctor because it takes lot of time for testing. Biopsy is done by removing skin tissues (skin cells) and that sample undergoes series of laboratory testing .There is possibility of spreading of disease into other part of body. It is more risky. Considering all the cases mentioned above, So Skin cancer detection using svm is proposed. This methodology uses digital image processing technique and SVM for classification. This technique has inspired the early detection of skin cancers, and requires no oil to be applied to your skin to achieve clear sharp images of your moles. In this way, it's quicker and cleaner approach. But, most importantly, due to its higher magnification, Skin Cancer Detection Using SVM can prevent the unnecessary excision of perfectly harmless moles and skin lesions.

A. Problems Statement

1. Skin Cancer: Skin cancer is the very common form of cancer in the United States. The skin cancers are divided into two major forms: Nonmelanomas and Melanomas. The Nonmelanomas usually form on the head, face, neck, hands, and arms. Another type of skin cancer is melanoma. It is more dangerous but less common. Melanoma occurs in the melanocytes, which are cells in the outer layer of human skin. The role of melanocytes produce the skin coluring calls as melanin, which gives skin its tan or brown color and helps protect the deeper layers of the skin from the harmful effects of the sun. Melanoma develops when melanocytes undergo malignant transformation .Melanoma may be cured if caught and treated early, but, if left untreated, the majority of melanomas eventually spread to other parts of the body. Early detection and surgery to remove the melanoma are successful in curing most cases of melanoma.

II. LITERATURE SURVEY

Skin cancer is the most common form of cancer, globally accounting for at least 40% of cases. It is especially common among people with light skin. The most common type of nonmelanoma skin cancer, which occurs in at least 2-3 million people per year. Of non-melanoma skin cancers, about 80% are basal cell cancers and 20% squamous cell cancers. Basal cell and squamous cell cancers rarely result in death. In 2003, it was estimated that 105,000 people would receive a diagnosis of melanoma and a further 33,000 would die from the disease worldwide. In the United States, there was cause of less than 0.1% of all cancer deaths.

1. Detection of Skin Cancer Using Image Processing Techniques by et. al. Chandrahasa M, Varun Vadigeriand Dixit Salecha suggested a method to detect ski cancer in early stages using smart phone application by analyzing properties of the cancer, Asymmetry, Border, Colour, Diameter and Expansion (ABCDE). These properties are analyzed using different image processing techniques like Grey scale conversion, segmentation, contour tracing and histogram analysis.
2. M.Chaithanya Krishna [2016]: This paper uses segmentation as various clustering technique, features can be extracted by using ABCD (Asymmetry Index Border Colour Index Diameter) method
3. Mariam A.Sheha,Mai S.Mabrouk, Amr Sharawy[2012]: This paper presents an method for melanoma diagnosis applied on a set of digital images. Features extracted by using gray level Co-occurrence matrix (GLCM) and Using Multilayer perceptron classifier (MLP) to classify between cancerous and noncancerous images
4. A.A.L.C. Amarathunga [2015]: This system used rule based and forward chaining approach to detect skin disease. Proposed system enables user to identify children skin diseases via online and provide useful medical suggestions. Used different data mining classification algorithms (AdaBoost, BayesNet, MLP and NaiveBayes) to predict and diagnose the skin disease. This only works for three skin diseases (Eczema, Impetigo and Melanoma)

 III. PROPOSED METHODOLOGY

Cancer image classification is an important task to generate classi?cation maps as no of world observation cancer increasing day by day and these cancer contains different tools capable of capturing imagery time to time and utilized for a wide range of application. Thus classification of cancer imagery has current area of researches and classification results can be used for different real-time application. This system proposed a novel approach for classification of six different classes’ actinic keratosis, Basel cell carcinoma, cherry nevus, dermatofibroma, Melanocytic nevus and Melanoma by utilizing Cancer imagery. To achieve an effective Cancer image classification framework this system isolates its works in various stage; these phases are important to give the better classification accuracy and the next page described these phases in details.

IV. METHODOLOGY SYSTEM

The methodology system was covered different step which shows a general step of diagnosis melanoma skin cancer. The block diagram shows the structure of the methodology system of this work as show in fig.1.

The first phase of the methodology was start with read the digital images and send it to the preprocessing stage in the skin cancer detection system.

V. PROPOSED TECHNIQUE

Most Automated Skin Lesion Diagnosis methods adopt the standard computer-aided diagnosis (CAD) pipeline which is illustrated in Fig: 1 and it consists of five general stages. After the image is acquired, it contains many artifacts such as hair and oil bubbles which could bias downstream processes are identified. Next, the lesion is segmented from the surrounding healthy skin.

After segmentation, discriminative features are extracted from the lesion. Features which are usually extracted are border, color, entropy, compactness, radial variance of the mask, coarseness. Finally; by extracting these features the detection is done which finally shows the risk probability of the lesion which is present in the image.

1. Image Pre-processing Techniques: The acquisition of the digital image of affected skin is the first and primary step in image processing. We are using images taken from commercially available digital camera or from Epiluminscence microscopy (ELM or Dermoscopy). Once image is acquired, then it goes for pre-processing. In first part of pre-processing digital images of skin cancer, collected in Bitmap or JPEG format from different sources are converted to indexed images. It converts the ordinary image to first RBG then grayscale and at the end binary. It makes an image suitable for a particular application. The second part of preprocessing involves enhancement of image (edge highlighting, sharpening, deblurring brightening, change in contrast, masking, hair removal, cropping or resizing and/or noise removal). For border detection of skin lesion we are using Canny Edge Detection technique.
2. Segmentation Techniques: Image segmentation involves image partitioning into multiple segments or regions of interest. It helps in to grouping similar characteristics regions. It is a process of extracting and representing information from the image to group pixels together with region of similarity. In this process a label is assigned to each pixel, such that pixels with same labels share common visual characteristics. Image segmentation is used to locate skin lesions and their boundaries. This paper includes Watershed segmentation, because of its popularity due to generation of less complex computational results. Steps involved in the watershed algorithm are. 1. Read the colour image and convert it into gray scale. 2. Using the gradient magnitude as a segmentation function. 3. Mark the fore ground objects.4.Compute the back grounds Markers.5.Tracing the exterior boundaries of the objects. 6. Performing the canny edge detection of the objects.
3.  Feature Extraction: In order to educate the masses to recognize melanoma in its early stages in 1985, group from New York University . devised the ABCD acronym (Asymmetry, Border irregularity, Colour variegation, Diameter > 6mm). It is one of the easiest guides to the most common signs of melanoma. Further, Stolz, W. established this diagnosis scheme for dermatoscopic images known as the ABCD rule of dermatoscopy. The characteristics needed to diagnose a melanoma as malignant are.

a. Asymmetry: To assess asymmetry, the melanocytic lesion is bisected by two 90º axes that were positioned to produce the lowest possible asymmetry score. If both axes dermocopically show asymmetric contours with regard to shape, colours and/or dermoscopic structures, the asymmetry score is 2.

 If there is asymmetry on one axis only, the score is 1. If asymmetry is absent with regard to both axes the score is 0.

b. Border Irregularity: The lesion is divided into eighths, and the pigment pattern is assessed. Within each one-eighth segment, a sharp, abrupt cutoff of pigment pattern at the periphery receives a score 1. In contrast, a gradual, indistinct cut-off within the segment receives a score of 0. Thus, the maximum border score is 8 and the minimum score is 0. In order to calculate border irregularity, there are different measures such as: compactness index, fractal index, edge abruptness, pigment transition. Compact index, fractal dimension, and edge abruptness has been calculated.

• Compact Index: Density index (compact index/CI) is the measurement of the most popular form of barrier which estimates unanimous 2D objects. However, this measure is very sensitive to noise along the boundary. This can be determined by using the following equation:

• Fractal Dimension: Fractal has characteristics self-similarity, and properties to the scale/size. Each section has a fractal which is different scale with the whole fractal. These characteristics causes suitable for fractal compression techniques. Another characteristic is fractal dimension. Dimension size generally an integer, such as the line has 1 dimension, the field has 2 dimensions, and cube has 3 dimensions and so on.

However, fractal dimension is a strange as it may worth fraction. This fractal dimension can be used as a characteristic of an image. Fractal dimension can be calculated by the method of calculation of the box (box-counting). To find the fractal dimension of an image, the Haussdorf dimension calculation method is simpler and effective one.

Consider the line with 2- dimension cube of side e and let N (e) is the smallest of esided cubes that can cover this line. The dimension of this line is then:

A. Advantages

1. We can easily detect and treat pre- cancerous condition in order to prevent skin cancer from developing.
2. Find skin cancer as early as possible in order to improve the chances of full recovery.

B. Disadvantages

1. It is not discorved and therefore not treated althoughit sgould be.
2. Sometimes unusual is discorved during screening and thought to be cancer. This may make you worry and feel anxious , and you need do to further tests.

VI. RESULTS

I have collected skin cancer images from internet. They were undergone various pre-processing techniques like gray scale conversion, median filter maximum entropy, ABCD method to classify cancerous and noncancerous image, output of above image would be ‘cancerous’ shows in fig belows-

Conclusion

It can be easily concluded that the proposed system of skin cancer detection can be implemented using gray level co-occurrence matrix and support vector machine to classify easily whether image is cancerous or non-cancerous. Accuracy of proposed system is 95%. It is painless and timeless process than biopsy method. It is more advantageous to patients.

References

[1] www.skincancer.org [2] Muhammad Ali Farooq, Muhammad Aatif Mobeen Azhar, \"Automatic Lesion Detection System (ALDS) for Skin Cancer Classification Using SVM and Neural Classifiers,\" 2016 IEEE 16th International Conferenc [10] [3] Dr. S.Gopinathan, S. Nancy Arokia Rani,\"The Melanoma Skin Cancer Detection and Feature Extraction through Image Processing Techniques,\" International Journal of Emerging Trends Technology in Computer Science (IJETTCS), Vol. 5, [11] [4] Rigel, D.S., Russak, J. and Friedman, R., “The Evolution of Melanoma Diagnosis: 25 Years Beyond the ABCDs,” CA Cancer Journal of Clinicians, No. 60, 2010, pp. 301-316. [12] Stolz, W., Riemann, A., C [5] Mariam A.Sheha ,”Automatic Detection of Melanoma Skin Cancer”, International Journal of Computer Applications, 2012. [10] [6] Anshu bharadwaj, “Support Vector Machine”, Indian Agriculture Statistics Research Institute. [11] [7] Maurya R, Surya K.S,\"GLCM and Multi Class Support Vector Machine based Automated Skin Cancer Classification, \"IEEE journal, vol 12, 2014.

Copyright

Copyright © 2022 Prof. S. G. Shinde, Shivani Nagnath Giram. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.