Category: Melanoma

Abstract —

Melanoma is the deadliest form of skin lesion which is a severe disease globally. Early detection of melanoma using medical images significantly increases the survival rate. However, the accurate recognition of melanoma is extremely challenging task. Since the joint use of image enhancement techniques and deep convolutional neural network (DCNN) has demonstrated proven success, the joint techniques could have discriminatory power on skin lesion diagnosis as well. To this hypothesis, we propose the aggregation algorithm for skin lesion diagnosis that utilizes a DCNN to extract the local features and classify medical images for melanoma disease. All experiments are performed using the data provided in International Skin Imaging Collaboration (ISIC) 2018 Skin Lesion Analysis towards Melanoma Detection. Experimental results show that our algorithm achieves excellent classification results for melanoma diagnosis.

Keywords — Skin Cancer, Deep learning, Convolutional Neural Networks, Deep learning, Melanoma Classification

I. Introduction

Yet if we mindfully practice protection against sun whole summer, it’s essential to maintain being observant about our skin in other seasons. During the year, we should consider our body skin from head to toe at least once a month, looking for any unusual lesions. To check if there is some extraordinary spot, we need to know what we’re looking for. As a general rule, to spot either melanomas or non-melanoma skin cancers, such as basal cell carcinoma and squamous cell carcinoma, we should notice any moles or growths, and any current growths that start to develop or change noticeably in any other way. The possible alarming signs are an itch or bleed.

Since melanoma is a very harmful skin disease, it is very crucial to catch melanoma. Melanoma is a type of cancer that most rapidly increases over the world. Example images for melanoma disease are provided by Figure 1. In fact, melanoma raises from the pigment cells in the human skin [1]. According to the annual report of the American Cancer Society (ACS) in the United States, 99,550 cases were diagnosed as new cases of melanoma, and the estimated deaths from this disease included up to 13,460 cases in 2018 [2]. Early detection and treatment of skin cancer may increase the survival rate of humans in the worldwide. Dermoscopy technology is developed to improve the diagnostic performance of skin cancer or melanoma. This technique is widely utilized to achieve a higher recognition accuracy rate of melanoma [3]. Unfortunately, the visual inspection and recognition by dermatologists is usually challenge due to the huge similarity among the different skin lesion types and it takes a long time, especially with a huge number of patients [3]. Recently, some studies have been proposed during the International Skin Imaging Collaboration (ISIC) Challenge to analyze the dermoscopy images seven classes which are Melanoma, Melanocytic Nevus, Basal Cell Carcinoma, Actinic Keratosis, Benign Keratosis Dermatofibroma, and Vascular Lesion.

In recent years, deep learning techniques have been widely acknowledged as the most powerful tool for image classification, since various DCNNs, such as VGG-Net [4] and Res-Net [5], have won the Image-Net Challenge in recent years. However, it also has been widely criticized that DCNNs may suffer from over-fitting when the training dataset is not large enough [6]. Although a pre-trained DCNN model can transfer the image representation ability learned from large-scale datasets, such as Image-Net, to the generic visual recognition tasks, the rigid architectures of DCNNs limit the ability in dealing with images where objects have a large variation in shape, size, and clutter.

II. Related Work

This section provides a review of previous research on DCNN for pattern recognition, in particular medical image analysis.

Since the performance of the models which were built using deep learning accomplished promising results on various tasks, we believe that the CNN-based approach may tackle the issue effectively. Despite the fact that real power of CNNs have been discovered recently, applications of CNNs in medical image analysis can be traced to the 1990s, when they were used for computer-aided detection of microcalcifications in digital mammography [7]. With revival of CNN’s owing to the development of powerful GPU computing, the medical imaging literature has witnessed a new generation of computer-aided detection systems that show superior performance in many tasks including computer-aided detection of lymph nodes in CT images, automatic polyp detection in colonoscopy videos, and automatic detection of mitotic cells in histopathology images. Research and developments of CNNs in medical image analysis are not limited to only disease detection systems, however, CNN’s have recently been used for skin lesion classification problems, pancreas segmentation in CT images, multimodality isointense infant brain image segmentation, and neuronal membrane segmentation in electron microscopy images.

In this paper, we propose several image enhancement algorithms joined with DCNN for Skin Lesion Classification. Recently proposed, well-known architectures, such as ResNet50 v1 [10], Inception v3 [11], Xception [12], DenseNet 201 [13], and Inception-ResNet v2 [14], achieved fairly high performances on classification task. Thus, in order to evaluate the performance of our architecture, we will be comparing our work with respect to these models.

III. Proposed Approach

A. Image Enhancement

The training images for the task of classifying melanoma, which consists of 10015 images, is downloaded from ISIC 2018: “Skin Lesion Analysis towards Melanoma Detection” grand challenge datasets [8-9]. An example image for each disease category is demonstrated in Figure 2.

Fig. 2. Example images from Skin Lesion Image dataset.

There are seven possible disease categories:

1. Melanoma
2. Melanocytic Nevus
3. Basal Cell Carcinoma
4. Actinic keratosis / Bowen’s disease (intraepithelial carcinoma)
5. Benign keratosis (solar lentigo / seborrheic keratosis / lichen planus-like keratosis)
6. Dermatofibroma
7. Vascular Lesion.

However, we classify images containing melanoma versus non-melanoma groups. Due to ISIC Challenge, datasets comprised of dermoscopic skin lesions taken by different dermoscope and camera devices all over the world, it is important to perform pre-processing for color normalization and illumination.

Despite having a deep and robust CNN algorithm, high recognition cannot be achieved unless a huge number of training images are provided. This is because of the over-fitting problem, where a network trained with a small number of images cannot generalize well on new, unseen test data. To solve this problem, we increased the number of training images using some data augmentation techniques. We performed image enhancement for every input image by preprocessing the dataset with a color constancy algorithm, augment brightness, saturation, and contrasts by a random factor in the range [0.9, 1.1], and the pre-processed results are as shown in Figure 3.

Fig. 3. Results of applying enhancement techniques to images found in the dataset.

B. DCNNs for image classification.

We start the second step by training the multi-class classification problem on five DCNN models: ResNet50 v1, Inception v3, Xception, DenseNet 201, and Inception-ResNet v2. A global average pooling layer is added after the base network and followed by three fully connected layers including the output layer. We use softmax as the loss function. Each model is trained for 300 epochs. The loss and validation loss are converged after around 240 epochs. As we discovered during the image enhancement process, the classes of the dataset are extremely imbalanced. We use class weights to balance the dataset. The higher weights will be given to the samples in the class with small size and the lower weights will be given to the samples in the class with large size.

C. Novel CNN Model

On top of that, we have created a simple CNN structure for the classification of lesion images using some of the ideas fundamental to LeNet, which was applied to recognize hand-written characters. Considering the fact that we need a deeper network for this task we add more layers and the

Fig. 4. The architecture of the deep CNN for skin lesion disease classification. The network consists of nine layers: four convolutional layers (a parallelogram with solid lines), two max-pooling layers (a parallelogram with dotted lines), and three fully connected layers at the end. Blue numbers under parallelogram denote the number of filters, black numbers denote the widths and heights of the feature maps and red numbers denote the number of neurons in the fully-connected layers.

final version of our network consists of 9 layers, as shown in Figure 4. The network has four convolutional and two pooling layers for feature extraction, and three fully-connected layers, in the end, for classification.

The main modification we make to the LeNet approach in our CNN model is using the adaptive piecewise linear activation (APL) function [15], instead of using traditional activation functions in convolutional layers. The proposed architecture improves the performance of the network, despite having some additional features which increase the training time (the detailed experimental results are shown in Section IV). Average pooling or max pooling is most commonly used in deep learning models. The former uses the average activation value over a pooling region, whereas the latter selects the maximum activation value. The max pooling technique is used in our network. We apply the same method for all three pooling layers over a [image: image1.png] pixel window with stride 2.

D. Adaptive Piecewise Linear Units

The activation function [image: image2.png] of an APL unit [image: image3.png] is a sum of hinge-shaped functions,

The result of Eq. (1) is piecewise linear activation function which computes the output of each neuron in feature extraction layers. Here, S denotes the number of hinges, and it is set as a hyperparameter. Variables and for are learned using gradient descent during training. The variables control the slopes of the linear segment, while the variables determine the location of images [18].

Using APL units requires the additional number of parameters to be learned; this is computed by 2SM, where M denotes the total number of hidden units in the network. As the number of hidden layers (M) in our network is predefined, the proper value for S which defines the complexity of the activation function is determined during the validation process and it is explained in Section IV

E. Training.

Our training procedure was identical for each model. First, we removed the final fully connected layer of each pre-trained network and replaced it with a randomly initialized (Glorot Uniform) matrix. We penalized the fully connected layer with both an L2 and an L1 regularizer, each weighted at 0.01. We then optimized these regularization terms along with a categorical softmax cross entropy term weighted at 1.0 with the Adam optimizer. All weights were optimized at all steps of training. Our initial learning rate was 10−4 and when the validation accuracy did not improve for 10 steps, we scaled the learning rate by a factor of 0.1. The remaining parameters of the Adam algorithm were left at their default values.

Each of the models we optimized took input tensors with spatial sizes of either 299×299 (Inception, Xception, and Inception-ResNet) or 224 (Resnet and DenseNet). For evaluation and prediction, we first scaled the images to either 329 × 329 or 254 × 254, cropped the central region of the appropriate size, and fed it to each model’s respective preprocessing function. For training, we first scaled each image to normally distributed random dimensions centered at 30 more than the desired dimensions with a standard deviation of 15. We then randomly rotated the image using bilinear interpolation, randomly cropped a region, flipped vertically and or horizontally, and randomly adjusted the brightness and contrast with a max change of 30% in both cases.

F. Network Training

The network is trained using stochastic gradient descent (SGD) and all training-relevant hyper-parameters are set as given in Table 1

TABLE I. Hyperparameters for Training

• Hyperparameters
• Momentum
• Initial learning rate
• Learning rate decay coefficient
• Value

1. 0.9
2. 0.01
3. 0.005

We initialize the weights using Xavier initialization [17].

IV. Experiments and Results

All experiments are performed using the Keras [33] software package on the Ubuntu 16.04 operating system, running on a PC with Intel(R) Core(TM) i7-7700HQ CPU 2.80 GHz with an Nvidia GTX 1050 Ti GPU.

A. Experiments with Proposed Method

To demonstrate the advantages of using APL units over traditional ReLUs and tangent functions, we compare the results of our deep CNN network model with APL units in hidden layers versus ReLU functions and tangent functions. We only change the activation functions of our model to ReLU and tangent functions; other hyperparameters remain unchanged. We train all networks using the training dataset which consists of 10015 skin images and evaluate using a test set which has 1500 images. Evaluation results, given by Table 2, show that our network performs better with APL units in hidden layers than with ReLUs and tangent functions.

TABLE II. Evaluation of performance of our deep CNN with different activation functions.

• Activation functions
• Test accuracy
• With data augmentation
• Without data Augmentation
• ReLU

1. 85.25%
2. 84.36%

• Tangent

1. 82.76%
2. 81.15

• APL units

1. 88.20%
2. 86.42%

Moreover, our network with APL units reaches the highest training accuracy around 98 % after about 65 epochs, and then it remains stable. The training accuracy of the same network with ReLUs and tangent functions is not higher than 96 %, and they become stable after about 150 and 200 epochs respectively. This means that our network consumes fewer computing resources and works faster to learn features of fire and smoke when it has APL units in hidden layers than when it has ReLUs and tangent functions. The detailed training process of our network with different activation functions is illustrated in Figure 5. And confusion matrix for our CNN method is presented in Figure 6.

Fig. 5. The training curves of the deep CNN with different activation functions.

B. Experiments and results with resemble

At the time of writing our performance on the challenge hold-out set is unknown, but we report here our results on a held-out 10% of the provided training set. We evaluated our models after each 200 training steps using 20 random batches of class-balanced validation data. The top overall validation accuracy reached by each of the models during training is shown in Table 3.

C. Comparisons with other methods.

In order to show the efficiency, we compared our approach with other existed methods. We conducted experiments with four best-performed algorithms in the ISIC challenge [10], [19]. Experimental results are displayed in Table 4.

TABLE III. Validation accuracy for each model during training

1. Model
2. Test
3. Accuracy (%)
4. Our Model

• 88.2

1. Inception

• 87.7

1. Xception

• 86.2

1. Inception-ResNet

• 87.8

1. ResNet

• 86.2

1. DenseNet

• 85.7

Figure 6. Confusion matrix of our model. Visualization of error rates across skin lesion types.

TABLE IV. Comparisons with other methods

1. Methods
2. Image size
3. Accuracy (%)
4. Jinyi Zou

• 224
• 75.10

1. Shlomo Kashani

• 224
• 80.06

1. M. K. Amro

• 224
• 85.69

1. Our method

• 128
• 88.2

V. Conclusion

In this work, we show how it is possible to obtain a high classification accuracy through training medical images using multiple pre-trained models and a newly developed novel CNN model for 7-class skin lesion classification. Firstly, we performed several data enhancement techniques to improve the quality of the images. Then, we trained all images using different pre-trained models for classification and using our proposed network. The experiments show that with a carefully designed data augmentation scheme and transfer learning algorithm, the prevalent deep models pre-trained on natural images can be successfully trained for skin lesion diagnosis applications as well as training the CNN model from the scratch. In our experiments, we demonstrated that the latter idea outperformed transfer learning approaches.

Diseases of the Skeletal, Integumentary, and Nervous Systems

The human body is made up of nine systems at the organ system level. These nine systems are made up of organs that have certain structures that promote functioning together, and similar functions and the systems working together constitute an organism. Diseases that affect three of these nine systems, the skeletal system, the integumentary system, and the nervous system is the focus of this essay.

Skeletal System Disease: Osteoporosis

Epidemiology and Causes

Osteoporosis is a disease that weakens the bones due to a decrease in the density of bones, making bones fracture easily and more likely. Osteoporosis occurs as a result of inactivity, low calcium level (decrease in vitamin D or calcium intake), aging, and estrogen deficiency (low circulating sex hormones), causing bones to weaken. Certain disorders can also deplete the amounts of the components that assist in maintaining the density and strength of bone. The symptoms of osteoporosis may not be noticed until a bone fracture occurs. These fractures may occur with after a minor fall, or with little or no force applied, as opposed to when the same force is applied to healthy bones, and the force does not lead to the bone fracture. Diagnosis for people at risk of osteoporosis is done by testing their bone density. Osteoporosis can be prevented by managing risk factors through sufficient intake of adequate calcium and vitamin D, and engaging in weight-bearing exercises, and it can be treated by taking bisphosphonates or calcium supplements that improve Vitamin D levels in the bones of the body, and also doing exercises that include weight-bearing activity.

Bones contain calcium and phosphorus, these are minerals which make bones hard and dense, which is a necessary weight-bearing structure to support the human body. To maintain bone density for support and other functions, the bones of the body require an adequate supply of calcium, and an adequate supply of vitamin D from food diet and manufacturing in the skin using sunlight, in order to absorb calcium from food and integrate it into the bones. For bone maintenance, there must also be adequate production of the proper amounts of several hormones, such as parathyroid hormone, calcitonin, estrogen, and testosterone.

Bones are continually broken down by osteoclast cells and reformed, in order to adjust to the necessary support required of them, and changing demands placed on them. This process is called remodeling, and in this process, small areas of bone tissue are taken out continuously, and new bone tissue is also continually deposited. Remodeling, therefore, affects the shape of the bones and their density, making bones grow in width and in length as the body grows during adolescence, and sometimes expand in width later life, without growing longer. Bone formation rate in young adults is higher than the rate at which bones are broken down, therefore there is a progressive increase in the density of bones until about the age of 30 when the bones are at their strongest. As aging occurs, breakdown of bones exceeds its formation, and the density of bones slowly decreases, and if the body does not maintain adequate bone formation by an adequate supply of vitamin D, and exercising, bones will continue to lose density and may become fragile, and eventually result in osteoporosis

Types of Osteoporosis

There are two main types of osteoporosis, these are Primary osteoporosis and Secondary osteoporosis. Primary osteoporosis mostly occurs when bone loses its density, and it occurs spontaneously, while secondary osteoporosis is majorly caused by another disorder or by intake of a drug.

Primary osteoporosis

Major causes of primary osteoporosis include deficiency of estrogen, increasing the likelihood of bone breakdown and rapid bone loss, in men and women, with the effect more obvious in women, due to the rapid decrease that occurs during menopause, but low levels of male sex hormones also contribute to osteoporosis in men. Another cause of bone loss in primary osteoporosis is low levels of calcium intake or low levels of vitamin D, resulting in calcium deficiency, and thirdly, increased activity of the parathyroid glands leading to the glands to releasing too much parathyroid hormone, therefore stimulating bone breakdown. Other factors include the use of certain drugs, smoking, use of heavy alcohol, and a family history of osteoporosis.

Secondary osteoporosis

Secondary osteoporosis results from existing disorders, Examples of these disorders include chronic kidney diseases and hormonal disorders, for example, Cushing disease, hyperparathyroidism, and diabetes mellitus. Secondary osteoporosis could also be certain types of cancer, for example, multiple myeloma, and other diseases for example rheumatoid arthritis. Other causes of secondary osteoporosis include the use of drugs such as antiseizure drugs, chemotherapy drugs, and progesterone, excessive consumption of alcohol and caffeine, and cigarette smoking.

Symptoms and Prognosis of Osteoporosis

Osteoporosis might not be obvious at first, because bone density loss is gradual, and some people do not develop symptoms. Bones that tend to break (fracture) is an indication of osteoporosis, and fractures tend to heal slowly in people that develop osteoporosis, and this may lead to deformities, for example, curvature of the spine. When people have had bone breakage due to osteoporosis, they tend to be at a higher risk of having more of such fractures. Fractures that occur on the nose, collarbone, and ribs are, however not considered to be the result of osteoporosis.

Prevention and Treatment

Osteoporosis prevention is generally more successful than its treatment, as it is easier to prevent bone loss density than to restore bone density once it is lost. Prevention involves the management of risk factors, including avoiding excessive consumption of alcohol and caffeine, consuming adequate amounts of vitamin D, and engaging in weight-bearing exercises. People that cannot consume the daily recommendation of vitamin D from diet alone, would need to take supplements to make up for the deficiency. Other measures include the prevention of fractures. Treatments administered to patients that develop osteoporosis include ensuring that adequate intake of calcium and vitamin D which helps the body absorb calcium, is consumed, and engaging in exercises that will improve bone density. When treating people that develop osteoporosis, doctors administer certain drugs and manage the conditions that could worsen the condition. treatment is usually recommended.

Integumentary System Disease: Melanoma

Epidemiology and Causes

Melanoma is a skin cancer that has its origins from the pigment-producing cells in the skin, known as melanocytes. Melanoma might affect normal skin or begin in existing moles on the skin. They appear as irregular, raised, or flat patches on the skin with either spot of different coloration, or blackish gray lumps. In order to diagnose melanoma, doctors must do a biopsy. Melanomas are then removed surgically, but if the melanoma has already spread, chemotherapy drugs and radiation therapy are then used, but the complete cure is difficult. Melanocytes, which are the cells that produce pigment in the skin, produce melanin, the pigment that darkens the skin and gives the skin its distinctive color. Sunlight act as a stimulant to melanocytes, encouraging further production of more melanin but also increasing the risk of developing melanoma, which normally appears on normal skin like a new, small, pigmented growth, and this growth most often occurs on sun-exposed areas. Melanomas readily metastasize (spread) to other areas of the body, where it grows continually, destroying tissues.

Types of Melanomas

There are two most common types of melanomas, these are superficial spreading melanoma and nodular melanoma. The former accounts for about 70 percent of reported cases of melanoma and occurs mainly on women’s legs and men’s trunks while the latter accounts for about 30 percent, grow very rapidly, and appear anywhere on the body. People that had melanoma are more likely to develop new melanomas. Melanomas are not very common among dark-skinned people and when it does occur, it develops most often on the palms of the hands and soles of the feet, and in the nail beds. Melanomas are also very rare in children, except for congenital melanocytic nevus, which is a patch of skin that is dark-colored and present at birth, resembling a birthmark, which could lead to malignant melanoma if the size is big.

Symptoms and Prognosis

Melanomas vary in appearance and size, some are flat and irregular, and the brown areas contain small black spots, while others are swollen brown patches resembling keloids, and has spots that are colored red, white, black, or blue. Some melanomas look like a firm red, black, or gray lump. Amelanotic melanomas, which does not produce pigment may be pink, red, or slightly light brown. Developing a new mole, or changes in an existing mole, for example, enlargement with an irregular border, darkening of areas of the skin, inflammation and spotty color changes, bleeding, itching, tenderness, and pain may be warning signs of possible melanomas. If melanoma is suspected, then a biopsy is done, all or part of the growth is removed and the sample examined under a microscope, to determine whether the growth is melanoma. If the biopsy shows that the growth is a melanoma and the growth has not been completely removed, all the cancer must be removed.

Prevention and Treatment

Melanoma is mainly developed due to long-term exposure to sunlight, it therefore might be prevented, starting in early childhood, by avoiding the direct glare of the sun by seeking shade, and reducing outdoor activities, especially between 11 AM and 3 PM when the sun’s rays are normally strongest, wearing protective clothing and using sunscreen; at least sun protection factor (SPF) of 30 with UVA and UVB protection applied as directed, and reapplied every 2 hours, and skin examinations at least once a year, especially for people who have many moles on their bodies. Treatment involves removal of the tumors for people who have shallow melanomas (known as melanoma in situ), and immunotherapy, targeted therapy, which entails the use of drugs that attack a cancer cell’s inherent biological mechanism, or radiation therapy for tumors that have spread to localized areas. If melanoma has metastasized to further distant areas, however, surgery is generally not an option, mainly chemotherapy is used to treat melanomas that have spread, but the complete cure may not be possible.

Nervous System Disease: Multiple Sclerosis (MS)

Epidemiology and Causes

Multiple Sclerosis is a disease of the nervous system, the term refers to many areas of scarring (sclerosis) and it occurs when patches of myelin sheath (this is a substance covering more nerve fibers) and other nerve fibers found in the brain, optic nerves, and the spinal cord are either damaged or destroyed. This destruction is called demyelination. The cause of damage is not known but may likely be exposure early in life to viruses or unknown substances that somehow trigger the body’s immune system to attack its own tissues damaging the myelin sheath underlying nerve fibers, known as autoimmune reaction, this results in inflammation. Genes and environment may also have a role in multiple sclerosis, therefore having a parent or sibling with multiple sclerosis may increase the risk of developing the disease. Over time, multiple sclerosis gradually worsens, but at its onset, periods of good health alternate with bouts of sickness brought about by the disease. Experiences of people that have multiple sclerosis include vision problems and abnormal sensations, also weak and clumsy movements. Life span is usually not affected, except when the disorder is quite severe but over time, the brain may start to shrink in size due to axons being destroyed. Mostly, multiple sclerosis affects people within the ages of 20 and 40, but it can also affect those within the ages 15 and 60 years. It is known to be more common in women and is uncommon among children.

Patterns of Multiple Sclerosis

Multiple Sclerosis often progresses and regresses unpredictably, typical patterns of the disease include;

• Relapsing-remitting pattern: Relapses, defined by worsening symptoms occurring spontaneously, alternate with remissions defined by symptoms lessening or not worsening and lasting for months or years.
• Primary progressive pattern: Multiple Sclerosis progresses gradually with no obvious remissions or relapses, with temporary highs when the disease does not progress.
• Secondary progressive pattern: This pattern starts with relapses alternating with remissions much like the first pattern, followed by a gradual progression of multiple sclerosis.
• Progressive relapsing pattern: The disease progresses gradually, with sudden relapses interrupting progression.

Symptoms and Prognosis

Symptoms vary extensively for individuals, and from period to period in one person, depending on the set of nerve fibers that are demyelinated: if sensory nerve fibers that carry information become demyelinated, sensory symptoms result, causing problems with sensations. If motor nerve fibers sending signals to the muscles become demyelinated, motor symptoms result causing problems with movement. Early symptoms of multiple sclerosis include tingling and burning sensations, numbness, pain, reduction in sense of touch, and itching in certain parts of the body such as the arms, trunk, legs, or face. Other symptoms include dim or blurred vision, stiffness in leg or hand, and imbalance in walking causing gait to be affected. Dizziness, fatigue, and vertigo are also common in multiple sclerosis, and excessive warmth including warm weather, or a fever may temporarily worsen symptoms.

Later symptoms of multiple sclerosis involve shaky, irregular, movements, partial or complete paralysis, spasticity when weak muscles contract involuntarily, and painful cramps. All these symptoms may lead to interference in walking and general body movement in people, and people who cannot walk tend to develop osteoporosis, which is decreased bone density. Late symptoms also include slow, slurred speech, inability to control emotional responses leading to inappropriate laughter or crying, impaired thinking, and depression.

Multiple Sclerosis also affects the nerves controlling bowel movements and urination, leading to frequent and involuntary passage of urine, fecal incontinence, which is passing stool involuntarily, urinary retention, which is the inability to empty the bladder completely, and urinary tract infections due to the fact that retained urine is a breeding ground for bacteria.

Treatment

Treatment for multiple sclerosis is not uniformly effective, but ingestion of drugs that could prevent the immune system from attacking the myelin sheaths can be consumed. These drugs assist in reduction of the number of future relapses and some examples are Interferon-beta and Glatiramer acetate injections, Mitoxantrone, which is a chemotherapy drug, and Natalizumab which is an antibody that is given intravenously. These can help to reduce the frequency at which relapses occur and slow the progression of the disease. Corticosteroids are most often used for acute attacks. These work by suppressing the immune system, are administered for short periods to help in relieving the immediate symptoms that interfere with functioning, for example, loss of vision, or coordination. Corticosteroids, however, only shorten relapses and slows the progression of multiple sclerosis, but do not entirely stop the disease’s progression. Corticosteroids can also have numerous side effects, including increased susceptibility to infection, diabetes mellitus, weight gain that might lead to obesity, osteoporosis, fatigue, and ulcers, they are rarely used for a long time, but rather as needed. Other treatments include plasma exchange for severe relapses that corticosteroids cannot control and stem cell transplantation. Maintaining an active lifestyle by getting involved in regular exercises such as walking, swimming, or stretching could help people with multiple sclerosis, although they may tire easily, this also reduces spasticity and assists in the maintenance of cardiovascular and psychological health.

Physical therapy can help with balance maintenance, can help reduce spasticity and weakness of limbs, muscle tone, general movement, and the ability to walk. People with multiple sclerosis should be encouraged to walk on their own for as long as possible, as doing so will improve their quality of life and prevent depression. Avoiding extremely high temperatures, for example, hot baths or showers, and very warm weather might also help during the repression of multiple sclerosis as heat tends to worsen symptoms. In addition, people who become too weak and unable to move about easily may tend to develop pressure sores, extra care must be taken to prevent the sores from occurring, by constantly shifting positions and assisting them to stand or sit upright as often as possible. Disabled people can be rehabilitated with occupational, physical, and speech therapy.

Several diseases impact the body due to the genes inherited from parents, lifestyle, and exposure to harmful substances. Early detection of these diseases could assist in better chances of curing the disease in its early stage, prevention of related secondary symptoms, reducing the risk of rapid deterioration, and increased life span.

Abstract

Cancer treatments have been advancing over the previous decade because of the urgent need to identify new therapeutics. Melanoma accounts for only 4% of all dermatologic cancers, but it accounts for almost 80% of the deaths recorded from skin cancer because of its ability of deep invasion and spreading as well as Melanoma tumor cells were able to effectively block chemotherapy which arose the need for new Recent studies has shown that kinases play a crucial role in various types of cancers.

Protein kinases are key regulators of cell function That consist of one of the largest most functionally diverse gene families1. Kinases are enzymes that function by transferring a phosphate group to a protein while it removes another phosphate group from another protein (phosphorylation)2 using this mechanism kinases maintain cellular functions and complex functions such as the cell cycle. It’s crucial role in cell proliferation, survival, and migration when activated as well as many kinase mutations are associated with cancer initiation, promotion, progression, and recurrence. Which made kinase as the key target for developing cancer therapeutics using different kinase inhibitors.

Studies of oncogenic signaling pathways in human cancer has identified molecular targets for cancer therapeutics such as EGFR HER-2/Neu, JAK, Src, Rho, and many other. Kinase Inhibitors has shown a potent effect against cancer proliferation, survival, and metastases. Inhibiting Rho kinase which plays a major role in the organization of actin cytoskeleton and activates myosin II-dependent contraction of actomyosin which regulates cell-to-cell contact and cell migration using AT13148 and CCT12925 reduced metastasis and inhibited cell proliferation3. Inhibiting Src activity using PD180970 and PD166285 Inhibited c-SRC and Stat3 which lead to inhibition of melanoma cell growth and Bcl-XL and Mcl-1 expression4. Using organometallic kinase inhibitor for Mdm2 caused high expression of p53 which induced apoptosis. Using a combined therapy that targets kinases by using multiple kinase inhibitors would stop melanoma cells from spreading and proliferating.

Melanoma frequently metastasizes through lymph nodes4, making it really hard to treat using surgical treatment, chemotherapy, and radiotherapy because melanoma is resistant to systemic treatment. Which made it really important to identify molecular targets that are critical for melanoma cell growth and survival4. One of the most popular oncogenes is Src. Src family kinases (SFK) are involved in regulating a multitude of biological processes, including cell adhesion, migration, proliferation, and survival as well as that it is known to activate other oncogenic pathways. Targeting c-Src or Stat3 leads to inhibition of melanoma cell growth and Bcl-xL and Mcl-1 expression. Targeting Src may allow us to inhibit multiple oncogenic pathways leading to more anti-tumor effects.4

Stat3 signaling is a point of convergence for many tyrosine kinases, activation of Stat3 in tumor cell proliferation, and survival by enhancing expression of anti-apoptotic genes. Inhibiting both Src and Stat3 signaling pathways can lead to apoptosis. Studying Src and Stat3 pathway had shown that Src inhibitors disrupt the DNA binding activity and induce tumor cell death4. That was proven by couple steps first they proved that stat was constitutively activated in human melanoma cell lines a super shift assays were made with Stat1, Stat3, or Stat5 specific antibodies were performed using nuclear extracts of A2058 or JW melanoma cells which showed high activity for STAT3 in melanoma cells as well as tumor tissues showed an elevated Stat DNA-binding activity. To determine which tyrosine kinase signaling pathway is responsible for Stat3 activation in melanoma cells a panel of kinase inhibitors effect was analyzed Stat3 DNA-binding activity in melanoma cell lines. This panel included inhibitors for JAK family kinases (tyrphostin AG490), EGF-R kinase signaling inhibitor (PD158780), and Src inhibitors (PD166285, PD180970). AG490 was used to determine Stat3 DNA-binding which had little or no inhibition effect on Stat3 DNA-binding, as well as PD158780 didn’t show any inhibition. While PD166285 which has broad specificity for various tyrosine kinases including Src4, and PD180970 which is selective for Src and Bcr-Abl kinases showed a great inhibition of Stat3 DNA-binding. To confirm Src activity on Stat3 DNA-binding activity a western blot was performed using c-src antibodies and antibodies to detect phosphorylated src residue tyrosine 419 (Fig1). none of the inhibitors affected c-Src at the protein level but both inhibitors were able (PD166285, PD180970) reduce phosphorylated Src protein which suggests that activated Src kinase mediates constitutive Stat3 activation in human melanoma cells. Figure 1 Src is activated in the melanoma cells and treatment with either of the two Src inhibitors diminished the levels of activated Src in these cells.

Src inhibitors (PD166285, PD180970) blocked growth of human melanoma cells with elevated Stat3 activity, A2058 and JW melanoma cells with activated Stat3 were treated using (PD166285, PD180970) resulting in a dose-dependent growth inhibition and showed reduced Stat3 DNA-binding activity. Those two inhibitors were also able to induce apoptosis because it was able to down-regulate anti-apoptotic genes Bcl-xL, Mcl-1, c-Myc, and cyclin D1, the efficiency of the inhibitors was examined using Annexin V-PE staining analyzed by FACS as an early indicator of apoptosis4 fig2. Figure 2 Inhibiting Src kinase activity, but not EGF-R or JAK tyrosine kinases induces apoptosis of human melanoma cells that display Stat3 activity. (4)

This finding demonstrates that Stat3 is constitutively activated in a majority of melanoma cell lines and tumor tissues. As well as Src but not EGF-R or JAK tyrosine kinases is responsible for Stat3 activation. Concluding that by targeting Src which would disrupt the Stat3-DNA binding activity it would be able to inhibit tumor cells proliferation and metastasis by inducing apoptosis and disrupting signaling pathways.

Dasatinib is potent FDA-approved melanoma treatment, it is a SFK Inhibitor that blocks tumor cells from migrating to normal cells and invading in nmol concentrations without affecting proliferation and survival of cells. It targets majorly focal adhesion kinase and Crk-associated substrate5, down-regulates expression of matrix metalloproteinase-9 (MM-9), and Inhibits EphA2 RTKs which is overexpressed in melanoma cells5.

Dasatinib effect was investigated on eight cell lines that showed SFK activity measured by tyrosyl phosphorylation of their autophosphorylation site. SFKs showed a critical role in regulation of angiogenic factors such as interleukin-8 and vascular endothelial growth factor5. Inhibiting SFK/FAK signaling pathway completely abolished migration and invasion of melanoma cells. That was test by the publisher by first testing if Dasatinib was able to inhibit migration by seeding 1205 lu and A2058 into a 12 well cell culture allowing it to proliferate for 24h then scratching it and treating it with DMSO and Dasatinib, allowing it to rest for 24h then quantifying it by counting the number of cells that migrated into the wound. Cells treated with Dasatinib showed fewer number of migrations. Testing the ability of Dasatinib to down-regulate Expression of Matrix Metalloproteinase-9 (MMP-9has previously been identified as a downstream target of SFK/FAK/p130CAS signaling with a critical role in invasion) collected treated cells after 96 hours of treatment showed down-regulation of MMP-9 by using Biotrak activity assay system to analyze them as well as Dasatinib was able to block enzymatic activity of MMP-9. Dasatinib didn’t show any inhibition for cell proliferation and viability fig3. Western blot was made to analyze SFK and SFK and downstream substrates of SFKs, including FAK and Crk-associated CASsubstrate, p1305 using antibodies for autophosphorylation site in autophosphorylation site in c-Src (Tyr419 in human and Tyr416 in chicken) cross-react with the corresponding autophosphorylation sites in other SFKs. Dasatinib was able to block SFK autophosphorylation and trysol phosphorylation of SFK down Stream. The data presented previously proved that dasatinib it can inhibit signaling pathways that are involved in cell adhesion, migration, and invasion5. Figure 3Dasatinib does not inhibit proliferation and viability of melanoma cells.

Dasatinib and PD180970 inhibited SFK catalytic activity at low concentrations. Dasatinib was able to inhibit cell migration and invasion and PD180970 was able to induce apoptosis by down-regulating anti-apoptotic genes. Using both tyrosine kinase inhibitors as a combined treatment of melanoma might be efficient because of its ability to inhibit proliferation and metastasis at the same time.

Dasatinib was able to inhibit SFK and SFKs downstream but it wasn’t able to inhibit known growth factors and survival pathways for melanoma such as mitogen-activated protein kinases Erk1 and Erk2, AKT, p38, and Stat3 signaling. Inhibiting the Rho family kinases allowed3 inhibiting of Stat3 and other growth factors that inhibited in its role cell migration and was able to decrease cell proliferation in secondary host.

Rho family kinases Rock1and Rock2 belong to the AGC kinase family of serine/ threonine (PKA, AKT/PKB) plays a major role in regulation of cell migration and affects several components of the metastatic process including migration, local invasion, and cell proliferation. Rock signaling plays a significant role in stiffening the extracellular matrix which increases cell proliferation. Activation of Rho phosphorylates several substrates (LIMK1,2, and MLC2)3. Tumor cells can move in either an elongated, protrusive mode or in an amoeboid movement both depend on actomyosin contractile. CCT12954 and AT13148 are multi-AGC ATP-competitive AKT kinase inhibitors. Comparing those two recently designed inhibitors with Y27632, H1152, and CCT129254 showed there efficiency in blocking both the rounded ‘amoeboid’ and elongated, protrusive mode of cell migration suggesting that inhibition of Rock in melanoma could be beneficial in the treatment of metastasis.

[image: ]CCT129254 and AT13148 were tested against a panel of kinases, AT13148 (AGC kinase inhibitor ) had a potent activity against Rock1, Rock2, and AKT 1,2, and 3 while CCT129254 ROCK1 and ROCK2 and AKT2 in nmolar concentrations. Both Inhibitors showed low potency against MRCK (myotonic dystrophy kinase-related Cdc42-binding kinase -a and -b), MLC2 phosphorylation was inhibited as well as AKT downstream signaling as shown by the decrease in Thr246 phosphorylation of the proline-rich Akt-substrate was strongly inhibited. on-target efficiency of the compounds was further confirmed by the dose-dependent decrease in phosphorylation of LIMK-substrate cofilin, a well-established target of ROCK-dependent signaling. AT12148 was stable over a period of 24hwhile CCT129254 was shifted toward higher potency after 24h. Potency was determined by Rock inhibition and determining their 24-hour-EC50s. Inhibition of ROCK actomyosin contractility (essential for the maintenance of cell shape) by CCT129254 and AT13148 induced cytoskeletal collapse which reduces cell movement which mimic the action of complete inhibition of myosin II by blebbistatin. All those results shows that inhibition of ROCK using CCT129254 and AT13148 decrease melanoma cell motility. AT13148 was then rejected because of its high toxicity. CCT129254 treatment led to a strong decrease in the number of metastases, with a marginal [image: ]effect on the growth of the primary tumor (Fig. 4). it also affected the colonization of the lung parenchyma, survival in the bloodstream, and their extravasation that was tested by using luciferase-expressing 4599 melanoma cells that, after 40 hours of pretreatment with the compounds, were injected into the tail vein of athymic mice and their Lumi- nescence measured 24 days after the injection. (Fig. 5) shows that lungs injected with the CCT129254-treated cells exhibit a consistently lower luminescence count than the control or CCT130293-treated cell. Figure 4CCT129254 inhibits spontaneous melanoma metastasis, number of metastases into the lung after treatment

Figure 5Rock inhibition decrease growth of melanoma metastases when treated with CCT11299254

Since Dasatinib only affected metastases without showing any effect on tumor proliferation, I think it would be more efficient to use CCT129254 as a combined therapy with PD180970. PD180970 was proven to induce apoptosis by down-regulating anti-apoptotic genes in the primary tumor combining that with CCT129254 which decrease metastases and reduces the ability of the melanoma cells to efficiently grow in the lung. This combined therapy would have a great impact on primary tumor, metastases, and secondary host.

Inhibiting of MAP/extracellular signal-regulated kinase (ERK) kinase (MEK) sensitized melanoma cells to apoptosis induced by tumor necrosis factor-related apoptosis-inducing ligand as well as inhibition of MEK induces apoptosis of melanoma cells by mitochondrial [image: ]apoptotic pathway6 which is independent of other pathways and controlled by the up-regulation of proapoptotic BH3-only proteins PUMA and Bim and the down-regulation of the antiapoptotic Bcl-2 family member Mcl-16. U0126 was used to inhibit MEK which induced apoptosis, this was studied using a panel of melanoma cell lines (fresh melanoma isolates and [image: ]melanocytes) then treated with U0126 for 48h fig6 we can see that by inhibiting MEK6 using U0126 had induced apoptosis specifically in SK-MEL-28. To confirm the effect of U0126 MEK1 was silenced in two cell lines SK-MEL-28, and Mel-RM by transfecting the cells with MEK1 siRNA FIG7 which inhibited MEK1 when switched to culture and induced apoptosis but at the same time it wasn’t effective in multiple other melanoma cell lines shown in fig6. PD166285 and PD180970 were both proven to induce apoptosis by down-regulating anti-apoptotic genes. But didn’t show any [image: ]dramatic effects on the growth of SK-Mel-28 cells4 fig4. While inhibiting MEK using U0126 induced apoptosis in SK-Mel-28 cells6. Using U0126 combined with PD166285 and PD180970 would allow it to induce apoptosis in bigger range of melanoma cell types. Figure 6 induction of apoptosis by U0126 in a panel of melanoma cell lines, fresh melanoma isolates, and melanocytes.

Figure 7 Knockdown of MEK1 bysiRNA induces apoptosis of melanoma cells. a) western blot analysis of MEK1, pERK1/2, and ERK1/2 B)% of apoptotic cells after treatment

Figure 8 Photomicrographs of SK-MEl-28, A2058, JW after treatment for 24h with inhibitors

Treatment of metastatic melanoma by conventional chemotherapeutic and biological agents remains unsatisfactory. The recent cancer treatment development is focusing on targeting kinases. Kinases are directly related to cancer initiation and progression which make them a perfect target for cancer therapies. Using kinase inhibitors in cancer treatment showed a promising future and couple kinase inhibitors received FDA approval. Inhibiting FSK, STAT3, Rho family, and MEK as discussed previously have shown that they can be used in targeting cancer but inhibiting each one by itself had its limitations which we can overcome by using combined therapies of different kinase inhibitors.

References:

1. Protein Kinases: Introduction | CST https://www.cellsignal.com/contents/science/protein-kinases-introduction/kinases
2. Bhullar, K. S.; Lagarón, N. O.; McGowan, E. M.; Parmar, I.; Jha, A.; Hubbard, B. P.; Rupasinghe, H. P. V. Kinase-Targeted Cancer Therapies: Progress, Challenges and Future Directions. Mol Cancer 2018, 17 (1), 48. https://doi.org/10.1186/s12943-018-0804-2.
3. Sadok, A.; McCarthy, A.; Caldwell, J.; Collins, I.; Garrett, M. D.; Yeo, M.; Hooper, S.; Sahai, E.; Kuemper, S.; Mardakheh, F. K.; et al. Rho Kinase Inhibitors Block Melanoma Cell Migration and Inhibit Metastasis. Cancer Res 2015, 75 (11), 2272–2284. https://doi.org/10.1158/0008-5472.CAN-14-2156.
4. Niu, G.; Bowman, T.; Huang, M.; Shivers, S.; Reintgen, D.; Daud, A.; Chang, A.; Kraker, A.; Jove, R.; Yu, H. Roles of Activated Src and Stat3 Signaling in Melanoma Tumor Cell Growth. Oncogene 2002, 21 (46), 7001–7010. https://doi.org/10.1038/sj.onc.1205859.
5. Buettner, R.; Mesa, T.; Vultur, A.; Lee, F.; Jove, R. Inhibition of Src Family Kinases with Dasatinib Blocks Migration and Invasion of Human Melanoma Cells. Molecular Cancer Research 2008, 6 (11), 1766–1774. https://doi.org/10.1158/1541-7786.MCR-08-0169.
6. Wang, Y. F.; Jiang, C. C.; Kiejda, K. A.; Gillespie, S.; Zhang, X. D.; Hersey, P. Apoptosis Induction in Human Melanoma Cells by Inhibition of MEK Is Caspase-Independent and Mediated by the Bcl-2 Family Members PUMA, Bim, and Mcl-1. Clinical Cancer Research 2007, 13 (16), 4934–4942. https://doi.org/10.1158/1078-0432.CCR-07-0665.
7. Panka, D. J. Targeting the Mitogen-Activated Protein Kinase Pathway in the Treatment of Malignant Melanoma. Clinical Cancer Research 2006, 12 (7), 2371s–2375s. https://doi.org/10.1158/1078-0432.CCR-05-2539.

Abstract:

Automated melanoma recognition in using image processing technique from the available dermoscopic images in deep learning is difficult task because of the contrast and variation of melanoma in skin. It is mainly a non-invasive method so that it cannot contact with skin more forcefully. To overcome the disadvantages this paper proposes a method using very deep convolutional neural networks (CNNs).on comparing with older version of methods, in that lower-level features of CNN is employed. For more accurate classification in this method, we are using FCRN and CNN with the effective training limited data. Initially, Performance of Segmentation is done using residual networks using a image from the dataset followed by Classification by neural networks to check the abnormalities in skin. In this kind of classification technique, the network has more specified features from the segmented portion alone. The proposed technique is mainly evaluated on datasets and experimented with results that would show the performance in histogram and stages of melanoma which includes PSNR ratio.

Introduction:

Melanoma is actually a skin cancer, its initial stage is pigmenting on skin cells known as melanocytes. It is curable at the initial stage. It can be cured easily at starting stages. The major effect of this is from ultraviolet radiation with lower levels of pigmentation of skin. Lesions are mostly smaller in size and initially look like small dot.

Using the trained set of data, images get recognized and undergo FCRN which segments gradually and then using that enhanced image it will be introduced into classification to obtain the accurate results. Doing Pre-processing Techniques comprises of filtration, Enhancement. Performing Segmentation using FCRN and classification based on CNN. Showing any abnormalities while classification can be displayed at different levels.

Literature survey:

In [1] Deep Learning is implemented equipped with GPU (Graphics Processing Unit). Non-dermoscopic images which are taken from digital cameras which is referred to as clinical images. Clinical Images also known as non-dermoscopic images from which dermatologists recognize melanoma lesions and detect skin cancer. The non-dermoscopic images may contain noise effects and other factors such as hair so that they are eliminated by the method know as Pre-processing. Further, the Enhanced image or the output from the pre-processing unit is fed into CNN (Convolutional Neural Network). The advantage of this is diagnosis of skin cancer through Computer and it is a type of automated analysis. But it requires sophisticated algorithms. This is only for professionals and cannot help non-specialists. He concluded that cameras played a tool role in differentiating images. Here a complexity based on clinical images. As a result it increases accuracy and differentiation. In case of training a small data set is being used.

In [2] He says that idea on melanoma detection segments by 19-layer. CNN trained one end to other ends. In this medical segmentation, task is accurate. Advantageous of its effectiveness and efficiency. It is prone to suffer from variances. The results he arrived at Several effective training methods was implemented to overcome the challenges. Our approach methods when evaluating on an open challenge database of Skin Lesion Analysis Towards Melanoma Detection. This is vigorously arising to imaging with processing(both stages).

In [3]melanoma in automation recognition is part of deep leaning and which is derivation of CNN. Types of characteristics is being used for classifying. The main outcomes demonstrate high classification of very accurate compared with other datasets. The dataset contains hundreds of images. He concluded as accurate in classification is the best when compared with others.. These results are very promising and derived from CNN have high potential in melanoma recognition. In future work, we will creating a new model for Mat Conv Net.

In [4] RGB colour space is transformed is that whatever the colour information present in the images can be used for the purpose to differentiating the actual with affected. The most quantitative analysis on 250 images showed algorithm is efficient. in such a way the image intensity of each pixel is calculated through a weighted combination of the three RGB channels. This is effective for its speed function and flexible segmentation. One of the major disadvantageous is deformable models are normally semi-automatic the initial parameter values is used. This paper’s outcome is in this algorithm, it remains to be same unchanged. Further scopes for accurate shape and robust.

In [5] an easy and effective pre-processing for classification of melanoma by properties. the major part is aligned axis of tumour in simultaneous direction and (CNN) classified into the validation in 5-fold. The main advantage is to attain better in comparing with targeted property. Complexity is they require a minimum amount of trained datasets.

In [6] Chiranjeev Sagarl and Lalit gave their idea about melanoma as it is located in lesions which are detected. The melanoma detection can be done using colour spaces. With the help of colour spaces most important and necessary information regarding lesions which is embedded colour channel makes way for segmentation of a digital image. It is performed in clinical images. The image from mobile cameras does not have proper clarity The method used here is Colour channel-based Segmentation. Early diagnosis can cure melanoma in skin lesions and it is more accurate. Dermoscopic images not easy to get. Sophisticated algorithms are provided which is available only for professionals. This concludes as very simpler manner and easier manner on histogram. colour channels have been analyzed in different ways of the skin. The result is 94% accurate.

Proposed methodology:

Trained data are the images which act as an input for the process. The trained datasets are from ISIC where dermoscopic images is available. As image processing technique accepts trained images so that it can communicate easily with computer in software. The data sets are nothing but a skin lesion pigmented which may or may not be a melanoma. Median filter is used for filtration. It is used to eliminate noises present in the biomedical images. This is very simple and easy for implementing the smoother images. Also intensity is varied between one of the pixels. This is main in reduction of noise such as hair etc. so median filter normally reduces salt and pepper known as black and white noises that increase pixel and enhance the contrast and also increase intensity. Here, Adaptive Histogram Equalisation is helpful in increasing contrast of the images.

tage of the proposed FCRN is that it can make pixel-wise predictions, which is of valuable significance for skin lesion segmentation task.

Further FCRN improves the segmentation and gives the portion of affected lesion. The trained datasets results in effective way more accurately in the CNN. Convolutional Neural network is used for classification which gives good and accurate result compared to conventional methods. After the segmentation it preceeds over the classification process, here the images undergoes up sampling and down sampling process through deeper networks and gives the accurate result through histogram representation.

Results and discussion:

The main aim of the project is to recognize melanoma using Deep Learning Algorithm. Here Deep Learning Algorithm refers as a subclass of machine learning. Therefore FCRN (Fully Convolution Residual Networks). This is used in segmentation under deep layers for data representations. This experiment is carried on Intel Windows 7 MATLAB R2014b software. The Hierarchy starts from Pre-Processing methods, Segmentation followed by Classification of an input image. The dataset that was downloaded from the website: https://dataverse.harvard.edu.

This website provides ISIC melanoma skin images of analysing the tools for automation of segmentation and examination for the skin lesions with the aim of accurate melanoma detection from dermascopic images.

A particular image (.tif,.jpg,.png) is taken as an input for process from the trained dataset for testing. Since the dimensions of the image is not convinient so that it will adjust into the Command Window so it is resize the image in the workspace. In order to reduce to pixel in given image and the input image (RGB 24 bit) is converted into salt and pepper noise that is grey coloured image is (8 bit). It doesn’t have clarity in its appearance therefore, greyscale image is filtered using a Median filter which removes noise that results some visibility of pixels through our naked eye. With the filtered image, Contrast Enhancement is performed using Adaptive Histogram Equalisation that would enhance the interior portion of skin lesion and increased the contrast of the image.

Segmentation is performed using FCRN (Fully Convolutional Residual Networks) technique. The iterations goes upto 100 and enhanced for final segmentation. This segmented image is then classified using CNN (Convolutional Neural Network) based on abnormalities at different levels that is plotted in Histogram as final results.

CNN Output:

Histogram

Here, peak to signal ratio is measured from which it attains a peak value which is denoted by point.

Performance:

This contains training, validation, test and best levels at which the image is classified. Up to 4 epochs or duty cycles, the classification takes place. These epochs or duty cycles can be changed based on the requirements.

Command window output

It will denote the stages as early stage, Moderate and Severe based on the type of Melanoma images. PSNR value appears in the command window.

Further scope:

Further scopes of this include integrating probability distributions for complexity areas. Also, it enhances the differentiate the performance and improves this method on future applications.

Conclusion:

This paper proposes an image processing method for automated recognition which is based on very deep Convolution neural networks in order to proceed the recognition which consists of two steps: segmentation and classification. By framing two steps that without interaction of manually operated. On Comparing our method with other convolutions the CNNs can execute the data with higher differentiation through segmentation and classification. Further proceeding with a segmenting using FCRN gives an accurate result. Deep CNNs accompanied with trained sets which help to solve complexity in analysis of medical problems, with help of limited training data.. This will be more useful in easy detection. It is more useful in medical field for identifying skin cancer types at initial stage.

References:

1. Melanoma Detection by Analysis of Clinical Images Using Convolutional Neural Network -E. Nasr-Esfahani
2. Automatic Skin Lesion Segmentation Using Deep Fully Convolutional Networks with Jaccard Distance-Yading Yuan
3. Combining Deep Learning and Hand-Crafted Features for Skin Lesion Classification- Yading Yuan*, Ming Chao, and Yeh-Chi Lo
4. A Novel Approach to Segment Skin Lesions in Dermoscopic Images Based on a Deformable Model -Zhen Ma and João Manuel R. S. Tavares
5. Simple and Effective Pre-processing for Automated Melanoma Discrimination based on Cytological Findings -Takuya Yoshida

Abstract

Skin cancer is caused due to irregular growth of cells. There are three major types of skin cancer- Melanoma, Basal, and Squamous cell Carcinoma. Among the three the most dangerous form is Melanoma. If it is left untreated, it leads to the spreading of all over the body including organs. However, the early stage detection of Melanoma is really helpful to cure it. Computer vision can be used for the medical image diagnosis and it has been proved by many prevailing systems. This project focuses on implementing an interface-based computer-aided tool for the early detection of Melanoma skin cancer. The computer-aided tool makes use of non-dermoscopic images of lesions and it automatically extracts affected regions. It checks for Melanoma parameters like Asymmetry, Border, Colour, and Diameter (ABCD) for image segmentation. They are used to identify the Melanoma cancer segment. The simple input requirement and the robustness of its identification allow this Graphical User Interface-based Computer Aided tool to be an efficient tool within the diagnostic process of Melanoma.

Chapter1. Introduction

1.1 Introduction to Skin Cancer

Among all cancers known to us, Skin Cancer is the most common one. When the transformation and insignificant growth of normal cells take place, Cancer is said to have occurred. Tumor is the mass multiplication of cells. Tumor is also known as skin lesions. If malignant cells are present it is concluded as cancerous. This means that they also affect the neighboring tissues. Tumors may also travel through the bloodstream to organs. This process of spreading to other organs is known as metastasis. Tumors take all the nutrients required for the unaffected cells to work. Skin cancers are of three types: Basal Cell Carcinoma, Squamous Cell Carcinoma, and Melanoma. While only about 1% of diagnosed cases of skin cancer are Melanoma, they cause 75% of related deaths. It kills about 10,000 people each year, i.e., nearly one person dies of each Melanoma in every hour. According to World Health Organization currently, between 2 and 3 million non-melanoma skin cancers and 132,000 melanoma skin cancers occur globally each year. The global incidence of Melanoma continues to increase – however, the factors of cause of Melanoma are connected with exposure to the sun and a history of sunburn. These factors are within one’s own responsibility. If the diagnosis and treatment of Melanoma is done in its early stages, it can be cured. If diagnosis becomes late, it can grow deeper into the skin and spread all over the body. The main cause of Melanoma is exposure of skin to ultraviolet radiation.

Dermoscopy is a non-invasive examination technique based on the use of incident light and oil immersion to make possible the visual examination of subsurface structures of the skin[1]. Though the detection of melanoma using dermoscopy is higher than unaided observation-based detection, its diagnostic accuracy depends on the training of the dermatologist[2]. Even when the expert dermatologist uses the dermoscopy for diagnosis, the accuracy of melanoma diagnosis is estimated to be about 75-84%[3]. The computer-aided diagnostics is very helpful for its accuracy. Computer is not more intelligent than humans but it may be able to extract some information, like color variation, asymmetry, texture features, that may not be perceived by human eyes.

The key steps in a Computer-Aided diagnosis of Melanoma are: image enhancement of a skin lesion, Preprocessing of skin lesion, Segmentation of the skin lesion from the skin region, extraction of features of the lesion blob, and feature classification. The first model is being implemented in Matlab and after with the help of GUI, diagnosis is carried out.

Chapter-2. Literature Survey

Many researchers have been working on the Computer vision approach for skin cancer detection. For segmentation of skin lesions in the input image, existing systems either use manual, semi-automatic, or fully automatic border detection methods. The features to perform skin lesion segmentation used in various papers are shape, colour, texture, and luminance. Many border detection methods are reported in the literature[4,5]. Some of the methods include histogram thresholding[6] global thresholding on optimized colour channels followed by morphological operations[7], and Hybrid thresholding[7]. In [4], the author has introduced an automated Global border-detection method in dermoscopy images based on color-space analysis and global histogram thresholding which exhibits high performance in detecting the borders of melanoma lesions. In [8], the authors have used the technique of dividing the input image into various clinically significant regions using the Euclidean distance transform for the extraction of color and texture features. The ABCD rule of dermoscopy[9], suggests that asymmetry is given the most prominent among the four features of asymmetry, border irregularity, colour, and diameter. A number of studies have been carried out on quantifying asymmetry in skin lesions. In Some techniques, the symmetry feature is calculated based on geometrical measurements on the whole lesion, e.g. symmetric distance and circularity[10].

Chapter3. Work done

3.1 Proposed model in Matlab

The proposed model in Matlab is as follows

• Load Image: In Matlab, a non-dermoscopic image is being loaded. Image may not be enhanced, it may contain noise and unwanted hair.
• Preprocessing: Values of blue channel of image are extracted since blue channel values lead to higher SNR values. Image enhancement and noise removal is also done in preprocessing. Image enhancement means increasing contrast of image by mapping of intensity value to new values. Noise removal is done by various filters such as Wiener filter, N-D filter, Median filter, and Averaging filter. Among all these filters Median filter is used for the extracted blue channel values of image. Now the image is enhanced and noise free.
• Segmentation: It is process of separating affected skin lesions from normal skin. K-means clustering-based image segmentation process is used among all the segmentation processes. It is process of partitioning image into parts or regions. This is based on characteristics of pixels in the image. At first, it is formed into clusters. Clusters mean grouping of data points together which are so similar. K represents number of centriods, which means centre of cluster.
• Post-processing: In this post-processing Morphological filter is used. Morphology is a broad set of image processing operations that process images based on shapes. At first, each pixel is being compared with the neighbor’s pixel and it is normalized accordingly. Rule which is used to process the pixels is dilation. In this value of output pixel is maximum value of all the pixel in the neighborhood.

Chapter-4. Summary

Table 1 Quantitative analysis of the different denoising filters

• Images
• Filters
• Signal-to-noise ratio (SNR)

Red Channel Image

1. Wiener filter
2. N-D filter
3. Median filter
4. Averaging filter

• 12.4449
• 10.5746
• 13.9916
• 12.5419

Green Channel Image

1. Wiener filter
2. N-D filter
3. Median filter
4. Averaging filter

• 14.4867
• 11.5672
• 15.6644
• 14.5197

Blue Channel Image

1. Wiener filter
2. N-D filter
3. Median filter
4. Averaging filter

• 15.9593
• 13.0643
• 16.5475
• 15.8369

Grayscale Image

1. Wiener filter
2. N-D filter
3. Median filter
4. Averaging filter

• 13.9013
• 11.3304
• 15.2304
• 14.0110

Results:

• Matlab:
• Image enhancement:
• Noise removal:
• Segmentation using K-means Clustering:
• Morphological Filtering:
• GUI:

Chapter-5. Future work

Future work includes following steps

• Features extraction from the morphologic image and classify it as melanoma or normal skin.
• Implementation of various algorithms for the segmentation and classification and choosing the best one.

Chapter-6 References

1. Jain, S. and Pise, N., 2015. Computer-aided melanoma skin cancer detection using image processing. Procedia Computer Science, 48, pp.735-740.
2. Bafounta ML, Beauchet A, Aegerter P, Saiag P. Is dermoscopy (chemiluminescence microscopy) useful for the diagnosis of melanoma? Results of a meta-analysis using techniques adapted to the evaluation of diagnostic tests. Arch Dermatol,137:13,43–50. 2001.
3. G.Argenziano, H. Soyer, S. Chimenti, R. Talamini, R. Corona, F. Sera, and M. Binder, Dermoscopy of pigmented skin lesions: Results of a consensus meeting via the Internet Journal of the American Academy of Dermatology, vol. 48, pp. 679–693, 2003.
4. R. Garnavi, Computer-aided diagnosis of melanoma, Ph.D. dissertation, University of Melbourne, Australia, 2011
5. M.E. Celebi, H. Iyatomi, G. Schaefer, and W. V. Stoecker, Lesion border detection in dermoscopy images Computerised Medical Imaging and Graphics, vol. 33, no. 2, pp. 148–153, 2009
6. H. Iyatomi, H. Oka, M. Saito, A. Miyake, M. Kimoto, J. Yamagami, S. Kobayashi, A. Tanikawa, M. Hagiwara, K. Ogawa, G. Argenziano, H. P. Soyer, and M. Tanaka, Quantitative assessment of tumour extraction from dermoscopy images and evaluation of computer-based extraction methods for an automatic melanoma diagnostic system Melanoma Research, vol. 16, no. 2, pp. 183–190, 2006
7. R. Garnavi, M. Aldeen, M. E. Celebi, A. Bhuiyan, C. Dolianitis, and G. Varigos, Automatic segmentation of dermoscopy images using histogram thresholding on optimal color channels International Journal of Medicine and Medical Sciences, vol. 1, no. 2, pp. 126–134, 2010
8. M. Celebi, H. Kingravi, B. Uddin, H. Iyatomi, Y. Aslandogan, W. Stoecker, and R. Moss, A methodological approach to the classification of dermoscopy images, Computerized Medical Imaging and Graphics, vol. 31, pp. 362–373, 2007
9. W. Stolz, A. Riemann, and A. Cognetta, ABCD rule of dermatoscopy: A new practical method for early recognition of malignant melanoma, European Journal of Dermatology, vol. 4, pp. 521–527, 1994
10. V. Ng, B. Fung, and T. Lee Determining the asymmetry of skin lesion with fuzzy borders Computers in Biology and Med., vol. 35, pp. 103– 120, 2005

Abstract—

Skin diseases such as Melanoma and Carcinoma are often quite hard to detect at an early stage and it is even harder to classify them separately. Recently, it is well known that the most dangerous form of skin cancer among the other types of skin cancer is melanoma because it is much more likely to spread to other parts of the body if not diagnosed and treated early. In order to classify these skin diseases, “Support Vector Machine (SVM)” a Machine Learning Algorithm can be used. In this paper, we propose a method to identify whether a given sample is affected with Melanoma or not. The steps involved in this study are collecting labeled data of images that are pre-processed, flattening those images and getting the pixel intensities of images into an array, appending all such arrays into a database, training the SVM with labeled data using a suitable kernel, and using the trained data to classify the samples successfully. The results show that the achieved accuracy of classification is about 90%. (Abstract)

Keywords—Melanoma, array, images, Carcinoma, classify, machine, algorithm (keywords)

I. Introduction

A. Background and Motivation

Skin diseases are one of those sets of diseases whose number has been largely increasing day by day. Only in India, about 200 million people suffer from one or the other forms of skin diseases. People often neglect skin diseases and do not take necessary treatment. This is especially seen in rural and economically backward areas due to many factors such as lack of awareness, poverty, and lack of resources, etc. this is even higher when it comes to the case of Melanoma skin cancer. It is reportedly found that about 132,000 melanoma skin cancers occur globally each year [1]. When the people tend to approach a physician, it is quite difficult for the physician in order to exactly detect the type of skin disease the patient is getting affected with. Especially when it comes to the diseases like Melanoma, it is quite hard to differentiate without any tests being conducted. In men, it’s often found on the skin on the head, on the neck, or between the shoulders and the hips while, in women, it’s often found on the skin on the lower legs or between the shoulders and the hips [2]. Besides SVM, another technique can also be used to classify among diseases. That is the classification using “Neural Networks” [3]. However, SVM is a better technique to classify than Neural Networks because they have a strong founding theory. SVMs reach the global optimum due to quadratic programming, they have no issue for choosing a proper number of parameters, Also, SVMs are less prone to overfitting and they need less memory to store the predictive model also yielding results that are more readable.

B. Contribution

Many research papers are published on using different algorithms to identify diseases but there is very limited research on using one particular method to classify two or more different diseases. Here in this study, we proposed an efficient technique in which the database of pre-processed images are trained and tested, and are classified using SVM, a machine learning-based algorithm to identify whether the skin lesion is benign or malignant. This will be very helpful in the diagnosis of the Melanoma skin cancer efficiently.

C. Related Work

In order to detect Melanoma various research works have been done in the fields of image processing. Some have used MATLAB to analyze and investigate the best formats to carry out the analysis. Pre-processing techniques such as Hair removal, centering the image, shading effect, vignette, and black-border cropping [4]. For segmentation, techniques such as Otsu’s Thresholding [5], color space transformation, Watershed algorithm, and c-means algorithm were used [6]. Works such as introducing the Image-based screening techniques to differentiate similar diseases and Multi-SVM classifiers were done in this field [7].

II. Methodology

Figure 1: The proposed Block Diagram

A. Pre-enhanced images

The statistical data is taken from an image database [8] that is verified to be prone to a particular disease. This database of images are pre-enhanced i.e.; they have undergone techniques such as hair removal, centering of the image, and softening.

Figure 2: Pre-enhanced images taken from verified website

B. Image flattening and image to array

The pre-enhanced images are converted into the RGB format and all the pixel values of the image in RGB format are converted into a 1-Dimensional array. The recorded pixel intensities are scaled in between (0, 1). Here, we took a 64*64-pixel format. So around 12288 values for each image have been recorded.

C. Database of image Intensities

All the 1-Dimensional arrays are arranged into the form of a database by appending all the image pixel intensities of all the images. By doing this, a database is created with all the pixel intensities of all the images.

Figure 3: Database of pixel intensities of all the images

As it is observed in figure 3, all the pixel intensities are taken and are formed into a database which is further then used to train the SVM as it is given as the input data.

D. Training the SVM

A Support Vector Machine is nothing but a machine learning algorithm which can classify among two or more classes [9], [10], [17]. The classification happens because of different kernels which are used as hyperplanes to differentiate among the classes [15]. The accuracy and precision of the SVM mainly depend on the Kernel used and the boundary values defined [4]. So, a suitable Kernel has to be taken in order to achieve better results. In this study, to differentiate among Melanoma and Non-Melanoma, a “Linear kernel” is used. Linear kernels are used when the classes which are to be separated don’t have many features in common [11]. The linear kernel is also one of the simplest of all the kernels available. When we want to classify two classes which are having more features in common, then other kernels such as Polynomial Kernels are to be used to achieve better accuracy and precision [12]. Gamma kernel is used to define the boundary values of the SVM [12]. The linear kernel’s equation is as follows:

This equation involves calculating the inner products of a new input vector (x) with all support vectors in training data [13]. The coefficients ‘y’ is the distance from the hyperplane to the feature and ‘c’ is an optional constant [13]. All the values, which were added into a database as mentioned, are given as input to the SVM and is trained to differentiate the classes as it is labeled data.

E. SVM classification

When unknown data is given to the SVM, it classifies the sample based on the training samples [10], [16]. Hence, SVM classifies the image whether it belongs to Melanoma or Non- Melanoma. As the “enum” function is used, the algorithm identifies the matched Melanoma samples as 0 and unmatched samples as 1.

III. Results

When working with SVM, The result depends on how well the SVM is getting trained. So, more the number of images the SVM is trained with, the higher is the accuracy. We tried training the SVM with different sets of images as input and we achieved the accuracy as follows:

• Images(Training)
• Precision
• Time(min)

1. 600
2. 70%
3. 9
4. 900
5. 80%
6. 13
7. 1400
8. 84%
9. 37
10. 1700
11. 90%
12. 46

Table 1: Table showing number of images, precision achieved, and time taken for the code to get compiled

As we can observe, with the increase in the number of images to be trained, the precision is increased also with an increase in the compilation time. In python, about 600 images data were approximately compiled in 9 min whereas in MATLAB, it took about 2 hours. This is the main reason for continuing this study in python.

• Image no.
• Predicted result

1. 1
2. 2
3. 1
4. 3
5. 4
6. 1
7. 5
8. 6
9. 1
10. 7
11. 8
12. 1

Table 2: Classification of Melanoma and Non-Melanoma

Table 2 shows the classification done by SVM as it identifies the samples belonging to Melanoma and Non-Melanoma. Here ‘0’ represents Melanoma and ‘1’ represents Non-Melanoma.

Figure 4: Image histograms of matched results

In figure 4, the image histograms of matched results are taken. It can be clearly observed that the RGB planes of both the images are very similar to each other thus proving that the images are malignant prone images.

Figure 5: Image histograms of unmatched results

IV. Conclusion and Future work

In this study, we presented that SVM can be effectively used to classify among the samples containing Melanoma and Non-Melanoma. It is observed that better results and precision can be achieved when the SVM is trained with more number of images.

The future work can be developing a product which can differentiate among two different types of cancers such as Melanoma and Carcinoma. Higher accuracy can be attained when other kernels such as Polynomial Kernels are used.

V. References

1. Skin cancers, World Health Organization, Available: [online] http://www.who.int/uv/faq/skincancer/en/index1.html Accessed December 2018.
2. Landis, Sarah H., et al. “Cancer Statistics, 1999.” CA: A Cancer Journal for Clinicians 49.1(1999):8-31
3. Codella, Noel, et al. ‘Deep learning, sparse coding, and SVM for melanoma recognition in dermoscopy images.’ International workshop on machine learning in medical imaging. Springer, Cham, 2015.
4. Mustafa, Suleiman, Ali Baba Dauda, and Mohammed Dauda. ‘Image processing and SVM classification for melanoma detection.’ 2017 International Conference on Computing Networking and Informatics (ICCNI). IEEE, 2017.
5. Vala, Hetal J., and Astha Baxi. ‘A review on Otsu image segmentation algorithm.’ International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)2.2 (2013): 387-389.
6. Manerkar, Mugdha S., et al. ‘Automated skin disease segmentation and classification using multi-class SVM classifier.’ (2016).
7. Manerkar, Mugdha S., et al. ‘CLASSIFICATION OF SKIN DISEASE USING MULTI SVM CLASSIFIER.’ 3rd International Conference on Electrical, Electronics, Engineering Trends, Communication, Optimization, and Sciences—2016.
8. Image dataset: https://www.kaggle.com/drscarlat/melanoma
9. Bono, Aldo, et al. “The ABCD system of melanoma detection.” Cancer 85.1 (1999): 72-77.
10. Alquran, Hiam, et al. ‘The melanoma skin cancer detection and classification using support vector machine.’ 2017 IEEE Jordan Conference on Applied Electrical Engineering and Computing Technologies (AEECT). IEEE, 2017.
11. Rosipal, Roman, Leonard J. Trejo, and Bryan Matthews. ‘Kernel PLS-SVC for linear and nonlinear classification.’ Proceedings of the 20th International Conference on Machine Learning (ICML-03). 2003.
12. Smits, Guido F., and Elizabeth M. Jordaan. ‘Improved SVM regression using mixtures of kernels.’ Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN’02 (Cat. No. 02CH37290). Vol. 3. IEEE, 2002.
13. Cesar Souza (2010, March 17) “http://crsouza.com/2010/03/17/kernel-functions-for-machine-learning-applications
14. Fatima, Ruksar, Mohammed Zafar Ali Khan, and K. P. Dhruve. ‘Computer-aided multi-parameter extraction system to aid early detection of skin cancer melanoma.’ International Journal of Computer Science and Network Security 12.10 (2012): 74-86.
15. Lv, Xiao. ‘A novel defect inspection approach using image processing and support vector machines in bolts.’ 2015 Seventh International Conference on Measuring Technology and Mechatronics Automation. IEEE, 2015.
16. Abe, Shigeo. Support vector machines for pattern classification. Vol. 2. London: Springer, 2005.
17. Dreiseitl, Stephan, et al. ‘A comparison of machine learning methods for the diagnosis of pigmented skin lesions.’ Journal of biomedical informatics 34.1 (2001): 28-36.

Melanoma Background

Melanoma in the early years was described as a benign mole that continued to grow and could be recognized by an irregular shape and discoloration of the mole (Sober et al., 1979). These cancer cells were well known for their ability to spread around the body quickly and avoid the therapeutic treatment, creating a dangerous situation (Li et al., 2001). When detected early, Melanoma could be removed successfully by a surgical procedure, but when metastasized, the cancer cells were much harder to kill, and treatment with chemotherapy and radiation were not much help, leading to a poor prognosis. (Sober et al., 1979; Irie et al., 2004). This led to copious amounts of research. Researchers linked that the first step in progression of Melanoma comes from a common melanocytic nevus, whose cells lost a checkpoint in the cell cycle and began to over proliferate (Clark et al., 1984; Carson et al., 2012). Inside these melanocytic nevi, over proliferation could be caused by a mutation in a proto-oncogene that activated the MAPK pathway, inhibited tumor suppressors, and allowed melanoma characterized cells to proliferate wildly (Florenes et at., 1998; Florenes et al., 1999; Kortylewski et al., 2001). Specifically, a high percentage of melanoma cases were linked to a mutation in the proto-oncogene BRAF (Colombino et al., 2012; Davies et al., 2002). Understanding this allowed scientist to tackle new ways of therapy. For example, immunotherapy and targeted therapy in melanoma cases were shown to be more successful than the previous therapies when studied and showed great room for more research (Deken et al., 2016; Moreno et al., 2015; Koya et al., 2012; Larkin et al., Long et al., 2014; Ribas et al., 2019; Robert et al., 2015) These findings struck a lot of interest in research that focused on the BRAF gene, its mutations, and its effects that lead to melanoma. By taking this knowledge about melanoma and emphasizing the focus on targeted therapies that will inhibit the effects caused by mutated BRAF in the MAPK pathway, the prognosis of Melanoma cancer will improve tremendously.

BRAF Gene

Characterization of the gene: BRAF

Improving the prognosis of melanoma and future treatments all starts by understanding this important gene, as it plays a lead role in a lot of actions. This proto-oncogene which encodes for serine/threonine kinase activates the MAPK signaling pathway by signaling to MEK which then sends a signal cascade down to ERK 1 and 2 (Kyriakis et al., 1992; Wajapeyee et al., 2008). The MAPK pathway is an important priority in understanding cancer cell proliferation and cancer cell survival (Kyriakis et al., 1992). Understanding the roll of activated MAPK brings hope for future therapeutics for melanoma patients (Cohen et. al., 2002). BRAF is located on chromosome 7 and is often mutated (Stagni et al., 2018). Studies have shown that when there are gains at chromosome 7 in the BRAF gene, there are high correlations of mutant melanoma diagnosis (Greshock et al., 2009; Helias-Rodzewicz et al., 2015; Stagni et al. 2018) A gene mutation in BRAF is found around 50% of the time in melanoma cells, and when this happens there is continuous activation of the kinase that starts the cascade signaling in cell proliferation (Curtin et al., 2005; Davies et al., 2002; Long et al., 2011; Si et al., 2018). With such a specific mutation linked to melanoma, there is prestigious information to be known about the BRAF mutation.

Mutations within BRAF can lead to melanoma

BRAF is most commonly mutated at the 600th amino acid, in which the amino acid Valine is switched to Glutamic Acid, giving the cell the newly mutated BRAF V600E gene (Davies et. al., 2002; Can et al, 2018). This gene may also mutate from Valine to Lysine at the 600th amino acid, which is called BRAF V600K (Wan et al., 2004). These mutations are very different from each other in ways, but both lead to big changes. As a result, these differences are best demonstrated by evaluation cohorts of Melanoma patients. For example, the proportion of melanoma cases that have a mutation in BRAF are more likely to have the V600E mutation (73%-76% of population) compared to the V600K mutation (19%-24% of population), and the V600K was mostly found in older patients compared to the younger patients with the V600E mutant gene. (Menzies et al., 2012; Ardakani et al. 2017; Bucheit et al., 2013). The mutations are also different in correlation with the prognosis and tumor characteristics, as the V600K mutant is found within patients with a short survival until stage IV and the tumor found in the neck and head area, compared to the V600E mutant found in longer survival times and in the areas such as limbs and trunk (Bucheit et al., 2013). This amino acid shift in either way (Valine to Glutamic acid or Valine to Lysine) leads to continuous signaling in the MAPK pathway, allowing cells in the pathway to have advantages when it comes to survival and making way for them to over proliferate (Can et al., 2018). This MAPK pathway being activated is an early event in the progression of melanoma (Cohen et al., 2002). Understanding this has led to therapeutic inhibition of proteins in the MAPK pathway.

References

1. Bucheit AD, Syklawer E, Jakcob JA, Bassett Jr RL, Curry JL, Gershewald. Davies MA. Clinical Characteristics and outcomes with specific BRAF and NRAS mutations in patients with metastatic melanoma. Cancer. 2013; 119(21): 3821-3829 doi:10.1002/cncr.28306
2. Can N, Tastekin E, Deniz Yalta T, Sut N, Korkmaz S, Usta U. Tozkir H. BRAF V600 Mutation Profile of Metastic Melanoma in the Thrace Region of Turkey. Turkish Journal of Pathology. 2018; 34(2):134-142 doi: 10.5146/tjpath.2018.01422
3. Carson C, Omolo B, Chou H, Zhou Y, Sambade MJ, Peters EC. Kaufmann WK. Aprognostic signature of defective p53-dependent G1 checkpoint function inmelanoma cell lines. Pigment Cell Melanoma. 2012; 25(4): 513-526 doi:10.1111/j.1755148X.2012.01010.x
4. Clark WH, Elder DE, Guerra D, Epstein MN, Greene MH, Van Horn M. A Study of Tumor Progression: The Precursor Lesions of Superficial Spreading and Nodular Melanoma.Human Pathology. 1984; 15(12): 1147-1165. Doi: 10.1016/ss0046 8177(84)80310
5. Cohen C, Zavala-Pompa A, Sequeria JH, Shoji M, Sexton DG, Cotsonis G. Arbiser J. Mitogen-activated Protein Kinase Activation Is and Early Even in Melanoma Progression. Clinical Cancer Research. 2002; 8: 3728-3733
6. Colombino M, Capone M, Lisboa A, Cossu A, Rubino C, De Giorgi V… Palmieri G. BRAF/NRAS Mutation frequencies amongprimary tumors and metastases in patients with melanoma. Journal of Clinical Oncology. 2012; 30(20): 2522-2529 doi: 10.1200/jco.2011.41.2452
7. Curtain JA, Fridlyand J, Kagoshima T, Patel HN, Busam KJ, Kutzner H. Bastian BC. Distanct set of genetic alterations in melanoma. N England J Med. 2005; 353: 2135-2147 doi: 10.1056/NEJMoa050092
8. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S. Futreal PA. Mutations of the BRAF gene human cancer. Nature. 2002; 417(6892): 949-954; doi: 10.1038/nature007766
9. Deken MA, Gadiot J, Jordanova ES, Lacroix R, van Gool M, Kroon P… Blank CU. Targeting the MAPK and PI3K pathways in combination with PD1 blockade in melanoma. OncoImmunology. 2016; 5(12) doi: 10.1080/2162402x.2016.1238557
10. Florenes VA, Lu C, Bhattacharya N, Rak J, Sheehan C, Slingerland JM, Kerbel RS. Interleukin 6 dependent induction of the cycling dependent kinase inhibitor p21WAF1/CIP1 is lost during progression of human malignant melanoma. Oncogene. 1999; 18(4): 1023-1032 doi: 10.1038/sj.onc.1202382
11. Florenes VA, Maelandsmo GM, Kerbel RS, Slingerland JM, Nesland JM, Holm R. Protein expression of the cell-cycle inhibitor p27Kip1 in malignant melanoma: inverse correlation with disease-free survival. The American Journal of Pathology. 1998; 153(1): 305-312 doi:10.1016/S0002-9440(10)655572-1
12. Greshock J, Nathanson K, Medina A, Ward MR, Herlyn M, Weber BL, Zaks TZ. Distict patterns of DNA copy number alterations associate with BRAF mutations in melanomas and melanoma derived cell lines. Genes Chromosome Cancer. 2009; 48(5): 419-428 doi:10.1002/gcc.20651
13. Helias-Rodzewicz Z, Funck-Bretano E, Baudoux L, Jung CK, Zimmerman U, Marin C… Emile JF. Variations of BRAF mutant allele percentages in melanomas. BMC Cancer. 2015; 15: 497 doi: 10,1186/s12885-015-1515-3
14. Homet Moreno B, Mok S, Comin-Andruix B, Hy-Lieskovan S, Ribas A. Combined treatment with dabrafenib and trametinib with immune-stimulating antibodies for BRAF mutant melanoma. OncoImmunology. 2015; 5(7) doi: 10.1080/2162402×2015.1052212
15. Irie RF, Ollila DW, O’Day S, Morton DL. Phase 1 pilot clinical trial of human IgM monoclonal antibody to ganglioside GM3 in patients with metastatic melanoma. Cancer Immunology. 2003; 53(2): 110-117 doi:10.1007/s00262-003-0436-1
16. Kortylewski M, Heinrich PC, Kauffman ME, Bohm M, Mackiewicz A, Behrnmann I. Mitogen- activated protein kinases control p27/Kip1 expression and growth of human melanoma cells. Biochemical Journal. 2003; 357(1): 297-303 doi:10.1042/0264 6021:3570297
17. Koya RC, Mok S, Otte N, Blacketor KJ, Comin-Anduix B, Tumed PC… Ribas A. BRAF Inhibitor Vemurafenib the Antitumor Activity of Adoptive Cell Immunotherapy. Cancer Res. 2012; 72(16): 3928-3937 doi: 10.1158/0008-5472.CAN-11-2837
18. Kyriakis JM, App H, Zhang X, Bangerjee P, Braughtigan DL, Rapp UR, Avruch J. RAF-1 Activates MAP kinase-kinase. Nature. 1992; 358(6385): 417-421 doi: 10.1038/358417a0
19. Larkin J, Ascierto PA, Dreno B, Atkinson V, Liszkay G, Maio M… Ribas A, Combined Vemurafenib and Cobimetinib in BRAF-mutated Melanoma. New England Journal of Medicine. 2014;371(20); 1867-1876 doi: 10.1056/nejmoa1408868
20. Li G, Schaider H, Satyamoorthy K, Hanakawa Y, Hashimoto K, Herlyn M. Downregulation of E-cadherin and Desmoglein 1 by autocrine hepatocyte growth factor during melanoma development. Oncogene. 2001; 20(56): 8125-8135 doi:10.1038/sj.onc.1205034
21. Long GV, Menzies AM, Nagrial AM, Haydn LE, Hamilton AL, Mann GJ… Kefford RF. Prognostic and CLinicopathological Associations of Oncogenic BRAF in Metastatic Melanoma. Journal of Clinical Oncology. 2011; 29(10); 1239-1246 doi: 10.1200/jco.2010.32.4327
22. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F, Larkin J…Flaherty K. Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma. New England Journal of Medicine. 2014; 371(20): 1877-1888 doi: 10.1056/nejmoa1406037
23. Menzies AM, Hayde LE, Visintin L, Carlino MS, Howle JR, Thompson… Long GV. Distingushing Clinicopathologic Features of Patients with V600E and V600K BRAF Mutant Metastatic Melanoma. Human Cancer Biology. 2012; 18(12):3242-3249 doi: 10.1158/1078-0432.CCR-12-0052
24. Mesbah Ardakani N, Leslie C, Greiu-lacopetta F, Lam WS, Budgeon C, Millward M, Amanuel B. Clinical and therapeutic implications of BRAF mutation heterogeneity in metastatic melanoma. Pigment Cell and Melanoma Research. 2017; 30(2): 233-242 doi:10.1111/pcmr.12569
25. Ribas A, Lawrence D, Atkinson V, Agarwal S, Miller Jr. WH, Carlino MS… Hamid O. Combined BRAF and MEK inhibition with PD-1 blockade immunotherapy in BRAF-mutant melanoma. Nature Medicine. 2019; 25(6): 936-940doi: 10.1038/s41591 019-0476-5
26. Robert C, Karaszewska B, Schachter J, Rutkowski P, Mackiewicz A, Stroiakovski D… Schadendorf D. Improved Overall Survival in Melanoma with Combined Dabrafenib and Trametinib. New England Journal of Medicine. (2015); 372(1): 30-39 doi: 10.1056/nejmoa1412690
27. Si L, Zhang X, Zhen X, Jiang Q, Bu L, Wang X… Guo J. Vemurafenib in Chinese patients with BRAFV600 mutation-positive unresectable or metastatic melanoma: an open label, multicenter phase 1 study. BMC Cancer. 2018; 18(1) doi:10.1186/s12885-018 4336-3
28. Sober AJ, Fitzpatrick TB, Mihm MC, Wise TG, Pearsons BJ, Clark WH, Knopf AW. Early Recognition of Cutaneous Melanoma. The Journal of the American Medical Association. 1979; 242(25): 2795-2799 doi: 10.1001/jama.1979.03300250051033
29. Stagni C, Zamuner C, Elefanti L, Zanin T, Del Bianco P, Sommariva A… Menin C. BRAF gene copy number and mutant allele frequency correlate with time and progression in metastatic melanoma patients treated with MAPK inhibitors. Molecular Cancer Therapeutics. 2018; 17(6): 1332-1340 doi: 10.1158/1535-7163.MCT-17-1124
30. Wajapeyee N, Serra RW, Zhu X, Mahalignam M, Green MR. Oncogenic BRAF Induces senescence and Apoptosis through Pathways Mediated by Secreted Protein IGFBP7. Cell. 2008; 132(3): 363-374 doi: 10.1016/jell.2007.12.032
31. Wan PTC, Garnett MJ, Roe MS, Lee S, Niculescu-Duvaz D, Good… Marais R. Mechanism of Activation of the RAF-ERK Signaling Pathway by Oncogenic Mutations of B-RAF. Cell. 2004; 116: 855-867 doi: 10.1016/s0092-8674(04)00215-6

Melanoma, otherwise known as malignant melanoma is the most dangerous skin cancer that we know of due to its potential to metastasize to other parts of the body including the liver, the lungs, bones and the brain. The most common patient groups are men over the age of 50, and women below the age of 30. The mechanism by which melanoma develops is due to damage to the skin DNA that goes unrepaired, triggering events that lead to rapid proliferation of melanocytes, cells that are located between the dermis and epidermis and produce the pigment melanin. The reason we chose melanoma is because we recently had dermatology, where we learned a lot about melanoma and its risk factors, prevention, etc. and being from Norway, we all are very aware of the high incidence in our country, and despite the eagerness of Norway’s population to look tan, there is not a lot of focus on prevention of melanoma.

Melanoma can originate from precursor lesions, or can appear as a new lesion entirely. The tumor will start with a radial growth phase during which metastasis and angiogenesis are not active processes. After a period of time, the tumor cells will start to grow vertically, invading deeper skin tissue. As with most cancers, early detection is the key prognostic factor. This can prove challenging as the distinction between melanoma and moles is known to be difficult. Patients are therefore asked to keep track of any changes in existing moles, or to be on the lookout for any new lesions that develop. During the growth phase, melanoma patients may notice typical signs such as changes in size, colour, irregularity, itching, bleeding. These changes typically occur on the back of men, or on the legs of women, however melanoma can occur anywhere on the skin as well as eyes or oral mucosa.

According to USPharmacist, “In 2015, 1.2 million Americans had melanoma and based on data for 2013 to 2015, 2.3% of Americans will be diagnosed with melanoma in their lifetime. The National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) Program estimates that, in 2018, melanoma will constitute 5.3% of all new cancer cases (91,270) and cause 9,320 deaths (1.5% of all cancer deaths).” One of the main risk factors for melanoma is overexposure to sunlight (UV rays), and this has generally due to sunbathing trends become an increased issue in various countries compared to earlier generations. People from more economically developed countries now tend to travel on vacations in order to enjoy the sun and “get a tan”. Suntans are currently looked upon as a sign of beauty, health and status. This has progressed into an increase in production and use of sun tanning beds in countries where there is less sun, such as in Scandinavia. Consequently, due to all these increased exposures to UV rays, the incidence and prevalence of malignant melanoma has steadily increased. When compared to earlier generations, upperclassmen would display their wealth through their pale skin, wearing large hats and resting in the shade, under parasols. Being in the sun, having darker skin was associated with outdoor labor and lower class/wealth.

The countries with the highest reported incidences of melanoma are New Zealand and Australia. Countries like Switzerland, Netherlands and Scandinavia also report among the highest incidences. The combination of cultural, socioeconomic and lifestyle factors all influence the development of melanoma in certain countries. In all these countries, fair skin – Fitzpatrick type I is the most common skin type found in the population. This skin type is associated with higher rates of melanoma as well.

According to a meta-analysis of SEER data, the 5-year survival rate exceeds 90% thanks to early detection. The 5-year survival rate drops to 60% if the melanoma has invaded lymph node tissue by the time of diagnosis. If the cancer spreads distantly, beyond the lymph nodes, the 5-year survival drops to less than 20%. In 1975, survival rates were 81.9% and have since, in 2012 improved to 93.2% for all races combined. For people with fair skin color, the rates increased from 82% to 93.2%. For people with darker skin colors, the rates increased from 56.7% to 69.1%. According to the same study: “The number of new melanoma cases was 22.8 per 100,000 men and women per year. The number of deaths was 2.6 per 100,000 men and women per year. Melanoma is the most common cancer in adults aged 20 to 30 years and the leading cause of cancer death in women aged 25 to 30 years. From 2003 to 2012, the melanoma death rate remained unchanged. The yearly incidence rate of melanoma was 21% higher in men than in women (1.7% vs. 1.4%), but the difference in incidence was less (13%) in whites. The incidence remained stable in Hispanic men, but it decreased by 1.7% every year in Hispanic women.” Diagnostic measures are very advanced, and usually not a problem in countries where the incidence is high. The problem rather lies with the attention and awareness of melanoma and its prevention.

The ABCDEs of melanoma:

• Asymmetry – is the lesion asymmetric?
• Border – is the border irregular?
• Colour – is the colour uniform? Are there multiple colours?
• Diameter/Difference – Is the lesion greater than 6mm?
• Evolution/Elevation – Has it changed over time?
• (F) – Firm to touch
• (G) – Growing
• Risk factors
• UV light exposure

A major risk factor of melanoma is exposure to ultraviolet rays. The main source of UV rays is from sunlight, but also tanning beds and sun lamps. The UV rays damage the cells of the skin – and when this damage affects the DNA of genes that control the skin cell growth, the skin cancer begins. The areas of the body that are more frequently exposed to sun are also the areas that are more prone to skin cancer. These areas are the face, neck and arms.

Moles

A mole, also called a nevus, is a benign pigmented tumor. They come in different size and shape and are usually harmless. But some nevi can develop to be malignant. In the dermatology course this year we learned that people with over ten nevi on the arm and fifty nevi on the body or more are at higher risk for developing melanoma. Humans are usually not born with moles, but if they are, the moles are called congenital melanocytic nevi. Risk of developing melanoma from this type is between 0-10%, and increases with size.

Another type is atypical moles or dysplastic nevi, they look a little like normal moles but have a tendency of malignancy. They are usually bigger and have an abnormal shape. Dysplastic nevi often runs in families and a small percentage mye develop into melanomas. An inherited condition with dysplastic nevi is called dysplastic nevus syndrome and this disease are associated with having a close relative with melanoma.

Skin & hair types

The risk of melanoma is much higher in people with pale skin, blond or red hair, green or blue eyes, or fair skin with freckles. That is to say that Caucasians are more prone to developing skin cancer.

Family history

If you have a first-degree relative with melanoma you are at higher risk of developing the disease, and around 10% of all people with melanoma have a family history of the type of cancer.

The increased risk might be because of a family tendency to have fair skin or some gene changes that runs in the family. Families also share the same lifestyle of being out in the sun.

Personal history

You are at a higher risk of developing melanoma if you already have had a melanoma. People who have had basal or squamous cell skin cancers are also at increased risk of getting melanoma.

Immunocompromised

If your immune system is weakened (from certain diseases or medical treatment), then your body can not work properly to fight against diseases, including cancer and melanoma.

Age & gender

Melanoma is more likely to occur in elderly, but can occur at any age. In fact, melanoma is one of the most common cancers in people younger than 30 years old, especially women. The risk is higher for women under the age of 50, but after the age of 50 the risk is higher in men.

Other diseases

Xeroderma Pigmentosum is a rare condition that affects the cells of the skin. It is an inherited disease that impairs the cells ability to repair damage to DNA. When having this disease you are at higher risk of developing melanoma.

Implementation of a health policy model

What is health policy?

According to World Health Organization, a health policy refers to a set of plans, decisions, and actions undertaken to achieve a health care goal within a society. It can achieve several things:

• It defines a vision for the future which in turn helps to establish targets and points of reference for short and medium term.
• Outlines priorities and the expected roles of different groups.
• It builds consensus and informs people.
• Choice of health policy

In Norway, the incidence of malignant melanoma is among the highest in the world, with about 1500 cases reported each year. Early diagnosis and appropriate surgical treatment cure many patients, as many as 80-90%. Therefore, the goal for the health policy should be prevention, early detection, diagnosis, and treatment of malignant melanoma, all achieved by increased attention to the risk factors and increased awareness in the general population.

In order to achieve this goal, we’ve chosen the ‘streams model’, also known as The Kingdon’s Three Stream Policy Window model. This model couples three streams, problem, policy, and political, to make a policy change. The problem, here the increased incidence of malignant melanoma in Norway and the lack of awareness to the risk factors. The policy stream is related to the alternatives that can be implemented and political stream is the willingness and ability of politicians to make a policy change.

We’ve decided not to use ‘the Stages model’ and ‘the Network model’ because we’ve found them to be inefficient. The former consist of negative feedback loops and re-alterations of the problem, and since our problem is already well-defined, the continuous re-alterations are unnecessary. The latter, ‘the Network model’,

Knowledge Exchange plan

With the incidences of melanoma increasing world-wide, it is necessary to take action. When it comes to Melanoma, a cancer which is fairly easily avoidable. The most important action that will be done is to inform the public. The most important aspects being the rise in occurrence, risk-factors, prognosis and of course prevention. But to get this knowledge to the public, we need a “Knowledge exchange plan”.

The message to share is straight forward, we want to raise awareness about Melanoma. Not just the disease in general, but maybe most importantly the risk-factors and prevention. The impact, or the action we want the “campaign” to have is to lower the occurrence of preventable incidences of Melanoma in the general public. This plan will need to be tailored to each of the identified audience respectfully. Especially when regarding what medium we use the distribute the information. We are proposing a campaign which will be subdivided into two parts. One part that targets the 12-25-year demographic and one more generalized part of the campaign.

With the increasing reach and influence the online blogger, YouTubers and “influencers” has been shown to have on the younger demographic, a campaign with a group of these people at the helm will have a great impact. We have many incidences in Norway where bloggers have raised their concerns about political issues which have led to greater awareness and change.

An example of this is from 2015 when the Norwegian blogger Sofie Elise Isachsen single handedly started a campaign to stop the use of palm oil in Norwegian food products. The campaign resulted in a nationwide boycott of all products including the oil and a few weeks later, these products were taken off the market. With the use of this kind of influence and reach, we could spread the awareness within the younger demographic. If we can reach the younger demographic and make it a part of their everyday conversations we think the campaign can make a difference.

The other part of the campaign would be a more generalized one, focused on the rest of the population. With more traditional modes of informing. Posters and billboards in the public room, infomercials, news articles and televised debates. We think that with this knowledge exchange plan, done right. We could make an impact on the preventable incidences of melanoma. Of course, this plan is not suitable for all nations. But for a country like Norway, where the online “influencers” have the reach that they do, we think this plan can make a difference. Say that the campaign we have proposed did make an impact and showed results, an important next step would be to remind the public regularly. A one time campaign is all well and good, but we think the campaign in some way or form should be repeated. We propose a 4 year interval between each campaign.

Current public health regulations

When it comes to regulations that the public health system in Norway are doing, or have done in the past, there is not a lot. There have been a few smaller campaigns consisting of televised infomercials. But they have not been followed by substantial results. This is why we think we need a new take on the campaigning that can really make a difference. As discussed earlier in the essay.

Public health challenges

As we see it, there is no issue with the diagnostic tools used, or the work done by the dermatologists that get the patients. The problem lies with the general awareness. If we can spread the awareness and make the public use sunscreen and do checkups with a dermatologist more frequently on their own accord, there will be a marked decrease in melanoma related deaths in Norway. The challenge will be to spread the awareness in a way that is understood and remembered.

Melanoma is an autosomal dominant disease which means that if you are to inherit it, the person before you would have had to have not one but two copies of the gene. If the melanoma is inherited it is called “familial malignant melanoma. If one of your parents had the two copies then you have a 50-50 chance of getting it. According to cancer.net, “Overall, about 8% of people newly diagnosed with melanoma have a first-degree relative with melanoma.” Also cancer.net says, “About 10% of people with melanoma have a family history of the disease. If a person has a close relative (parent, brother, sister, or child) who has been diagnosed with melanoma, his or her risk of developing melanoma is 2 to 3 times higher than the average risk.”When testing for melanoma, sometimes it can be as easy as looking at your skin, but the best way to diagnose it is to do a biopsy. The doctors do this by simply removing a part or all the mole and analyzing it. There are three different types of biopsies: punch, excisional, and incisional. A punch biopsy is a circular tool that they use to press into the mole to take a part of it out. An excisional biopsy and incisional biopsy are the same thing except that the excisional biopsy takes the whole mole while the incisional takes part of it.

Evolution

At a younger age, females have a greater chance of getting melanoma while men have a higher chance as they get older. According to the Skin Cancer Foundation, “From age 50 on, significantly more men develop melanoma than women. The majority of people who develop melanoma are white men over age 55. But until age 49, significantly more white women develop melanoma than white men (one in 150 women versus one in 215 men). Overall, one in 27 white men and one in 40 white women will develop melanoma in their lifetime.” While women can develop melanoma at younger age men are twice as likely to dies from melanoma at any age. Researchers aren’t exactly sure why men have a greater chance of getting it but some say it’s due to them not knowing as much. Several different surveys have concluded that women know more about melanoma than men do. Women also tend to apply sunscreen more than men as well as wear cosmetics which tend to have SPF in it. Although this is all a pretty accurate explanation, this can’t account for all men because women also use tanning beds and those who use them have a 34% increased risk of getting melanoma than those who don’t. The Skin Cancer Foundation says, “The estimated five-year melanoma survival rate for black patients is only 65 percent, versus 91 percent for whites.” This can be due to the black community not being able to get the proper medical treatments or checkups needed. Although this is true, skin cancer accounts for 1 to 2 percent of all cancers in blacks.