Melanoma - skin cancer reviewed

Lesions on various skin places used to be a sign of beauty. But until they are benign. As known about 50% of melanoma evolutes from benign lesions(congestion of melanocytes), other appear from unnatural growth of melanocyte cells. So both cases are based on melanocytes where they have invaded an epidermal and dermal layers of skin. There are normally about 10 to 40 lesions on human body. Some of them are native, inborn, some of them appear during 35 – 40 years. Later some of them may disappear.

Melanoma is not only a skin disease as it can also appear in eyes, respiratory system, even on cortex of brain. But most of cases (~90%) appears on skin.

As mentioned earlier it is hard to detect melanoma in early stages as it may appear as benign lesion. So if there is even a little suspicion about malignant lesions, it is necessary to attend screening at dermatologic specialist who usually photographs suspicious lesions and then analyses according to ABCD rule. Other non skin cases are even harder to detect. For instance if melanoma appears in lungs patient may have difficulties in breathing – like astma. Melanoma in bones may appear as bone ake.

Skin melanomas may be classified as follows:

• Radial melanoma (about 70% of skin melanoma cases) – appears from displastic lesions. Melanoma progress takes up to 5 years. Can be detected by using ABCD rule. More specific for elder ages. In early phases melanomas spreads in upper layer of skin – epidermis. Later it enters vertical growth phase after then cancer becomes dangerous as it starts to invade inner tissues;

• Nodular melanoma (about 15% of cases) – This type of melanoma rises very rapidly and is the most aggressive type of skin cancer. From the beginning this type of melanoma grows vertically up and down. So there is a danger from the beginning that melanoma will spread in to inner tissues. This type of melanoma appears to be dark black, blue, grey or red with smooth borders.

• Lentigoous maligna melanoma – appears usually in head and neck area bigger than 3 centimeters non symmetrical lesion. It takes long time to progress – up to 20 years. Most cases at 70 years old.

• Acral lentigous melanoma – more common for dar skinned people. It appears on palms of hands and foots – especially under nail of first finger. It takes short time to progress from 3 to 36 moths. Most common among 60 year age.

• Unclassified – other up to 5% of cases.

Generally speaking all melanomas growths in two phases: horizontal-radial and vertical. In radial phase lesion spreads in epidermis and papillary dermis without metastases. In this period patient can be easily cured. When growth enters vertical phase lesions spreads in to deeper tissues through dermis and deeper layers of skin. There begins metastases. In meny cases both phases come together. Most common metastases of melanoma are: in Dermis(50 -75%), Lungs(07 – 87%), Liver(54-77%), Brain(36 – 54%), Bones(23 – 49%), Digestive system(26 – 58%), Kidney(35 – 48%), Heart(40 – 45%).

How metastases are related to lesion size.

• <0.75mm – regional metastases(2-3%), far metastases(2-3%);

• 0.76-1.5mm – regional metastases(25%), far metastases(8%);

• 1.51-4mm – regional metastases(57%), far metastases(15%);

• >4mm – regional metastases(62%), far metastases(72%);

One of most important criteria is a number of metastases in lymphatic system. Depending on number of metastasis there is a statistics on survival from melanoma.

• No metastases in lymph – 73% to survive 5 years;

• 1 - 3 metastases in lymph - 55% to live 5 more years;

• 4 metastases in lymph - 26% to live 5 more years;

• macro metastases - 21% to life 5 more years and 12% to live 10 more years.

So number of metastases is proportional to survival.

When you should pay more attention to you skin lesions.

• Elder age people have to pay more attention as during age melanoma prognosis become more critical;

• Gender – man have more chances to get spreaded melanoma;

• Number of lesions on body – if there are more than average number of lesions on body;

• Skin color – white skinned people tend to have melanoma more offen than dark skinned people.

Main and only effective cure for melanoma stil is surgery when it is removed. But it is effective if there is no metastases. Otherwise complex treatment is taken without big chances. So it is important to diagnose melanoma in early stages before it has spreaded to other tissues.

References:

http://www.dermatologychannel.net/skincancer/melanoma/types.shtml

http://www.odosnavikai.lt/

Qualitatively high-quality, evenly and well illuminated top illumination-microscopic pictures of the lesions is the most important condition for the operability of this software.

If quality skin model is constructed, then recognizing skin cancer symptoms can be easier as there are many factors showing about threat of skin cancer. Of course this can’t give 100% results, as there are many shortcomings connected with skin lesion variety and interpretation errors. But some guides may help.

There are 3 main factors that can indicate risk of skin cancer. Recognizing skin cancer symptoms can be based on them. They are:

• Melanin presence in papillary dermis;
• Thickness of papillary dermis;
• Blood behavior around the lesion and inside it.

Firs of all Melanin presence in dermis. This is the main factor in recognizing skin cancer symptoms. If there is melanin spreaded in papillary dermis or even dermis, this is a big probability of being skin cancer symptoms, but not always. There are several sub factors in this issue like melanin spreading figure, depth, and melanin density within this shape. If there are more irregularities in spreading area there are more risks.

Other factor in recognizing skin cancer symptoms is papillary layer thickness. In not going in to deep too much there can be said, that thinner this layer, the bigger risk.

 

SkinSeg is simple tool used for skin lesion segmentation. This little program was developed by Intelligent Systems laboratory students: L. Xu, M. Jackowski, A. Goshtasby, C. Yu, D. Roseman, S. Bines, A. Dhawan, A. Huntley. Their method is working similar as in my earlier experiment with matlab pigmented lesion boundary tracing algorithm. First image is converted to intensity image and then the lesion edges are detected.

ant test results:

More informative description you can find here 

Program can be downloaded from here: http://www.cs.wright.edu/people/faculty/agoshtas/skinseg.zip 

This version of the Skin Cancer Segmentation program (skinseg) runs on the Windows 95/NT platforms. Make sure all files reside in the same directory after extraction. No setup program is required to install skinseg on your machine. To run, execute the program skinseg.exe.

DullRasor uses image processing techniques to analyze and segmentates skin areas with dark hair. This program removes dark hairs form images, and makes skin lesion images clean to further processing.

Many skin images contain various numbers of hairs. Other skin segmentation programs may mislead because of hairs – especially dark ones. One solution can be shaving skin before taking pictures of it. But shaving of skin adds more time to processing and this is uncomfortable and in some cases unesthetical.  Hence, a software approach for dark thick hair removal from skin images is needed.

There is only one program window:

DullRazor performs the following steps:

1. It identifies the dark hair locations by a generalized grayscale morphological closing operation,
2. It verifies the shape of the hair pixels as thin and long structure, and replace the verified pixels by a bilinear interpolation;
3. It smoothes the replaced hair pixels with an adaptive median filter.

The algorithm has been implemented in C on a SunOS 4.x workstation. (The program can be run on Sun Solaris workstations as well.) It has been tested on real nevi images with satisfactory results.

Download the Unix version of DullRazor (dullrazor.zip, 87KB).
Download the Windows version of DullRazor (dullrazor_wins.zip, 327KB.

My personal test on one of images:

|

|

V 

Read more in: http://www.derm.ubc.ca/dull_razor/

This is not completed list

The goal of any imaging methodology used in dermatology is to diagnose melanoma in early stages, because on it depends effectiveness of treatment. Investigations shows, that early diagnosis is more than 90% curable and late is less than 50% [1]. The diagnosis and successful treatment is often supplemented with permanent monitoring of suspicious skin lesions.

Doctor’s diagnosis is reliable, but this procedure takes lots of time, efforts. These routines can be automated. It could save lots of doctor’s time and could help to diagnose more accurate. Besides using computerised means there are good opportunity to store information with diagnostic information in order to use it for further investigations or creation of new methods of diagnosis.

Skin lesion imaging methods

We found that there are number of various imaging methods of skin lesions [2]. The simplest visualisation method is photography. This method gives only top layer skin image. In order to get deeper layer image there is oil immersion used. It reduces reflections of surface and brightens the image of epidermis – the second skin layer.

The different illumination method called epiliuminescence can be used in order to get the image from deeper skin layers. The light is directed straight in to these layers and reflected goes back through lesion giving more information about consistence of light absorbers in these layers. This method of illumination improves diagnosis accuracy up to 10 percent [2].

Other interesting solution of getting more information from skin is using multi spectral photography [2]. There is used narrow frequency band of light illumination. Those images give information about consistence and concentration of absorbers and reflectors in the skin. The idea is that different pigment of skin absorbs different light wave, determining the colour of our skin. When those photos are made with range of light waves, we can calculate the reflectance frequency characteristics of skin. And comparing to normal skin characteristic there can be made diagnostic decisions about skin pigment consistency.

Other imaging method using laser is called CLSM (confocal scanning laser microscopy). It uses red or near infrared low power laser beam to scan skin surface. This beam can be focused in to different deep to get the image of it. The deep is limited to 300µm, because of absorbance [2]. The distance between two layers (axial resolution) can be about 2 – 5 µm. The main disadvantage of this method is complicated acquisition of image from reflected laser beam.

Ultrasound visualisation is usually used to measure depth of melanoma [1]. The other uses of ultrasound are limited by very little tissue differences between normal skin and lesion. If there is no melanoma practically there is no any differences. When doctor diagnoses melanoma, then he uses high frequency ultrasound (over 30 MHz) to measure penetration depth in order to make correct cut during surgery.

In optical coherence tomography is used short near infrared light pulses focused to papillary dermis [2]. Reflected light is combined with reference light that is reflected from mirror system in order to determine the depth of papillary dermis. Measurement of the interference pattern allows determination of the position within the tissue where the light was reflected. Using recent technologies with ultra sort light pulse, the maximum obtained resolution is 2 – 4 µm. Visualisation depth is 1 – 1.5 mm.

The wide variety of methods shows, that there is no best universal visualisation. Some of them are used for different needs, other are very expensive. The choice of method depends on what features of skin lesions is wanted to visualise, and on availability of resources.

Algorithms of skin image processing

Digital dermatoscopic images itself does not provide formal and determined information. To get diagnostic information the digital image processing is used. Commonly used methods are based on geometrical feature extraction from image with lesion. The USA national health institute offers ABCD rule for classification of dermatological images in to benign, suspicious and melanoma [3]. ABCD are the letters of first feature words: A (asymmetry), B (border), C (bolour), D (dermatoscopic structures). According to these four values there are total dermatoscopic value calculated by formula:

N – Not known
Most of those software solutions have some major disadvantages. Firs of all, they are analysing geometrical and structural or colour changes extracted from plain surface picture, which does not contain enough useful information. Analysis is not connected with physical nature of skin image. There are no associations with histological parameters of skin.

Other disadvantage is that diagnostic results do not always support diagnosis decisions. If there is support of databases, then those pictures with diagnostic results are stored and should be used in decisions support algorithms. In now days there is tendency to request past data and knowledge to gain better diagnosis results in future.

Digital dermatoscopic images firstly have to be parameterised for automatic classification. The deep study of skin nature has to be done before to parameterise it. For extraction skin or lesion optical features it is very useful to use multi layer skin model [5].

The most common is four-layer skin model: Stratum Cornea, Epidermis, Papillary dermis and Reticular dermis.

Stratum Corneum is top thin layer, which is a protective layer consisting of keratin-impregnated cells and it varies considerably in thickness. Apart from scattering the light, it is optically neutral.

The epidermis is largely composed of connective tissue. It also contains the melanin producing cells, the melanocytes, and their product, melanin. In this layer there is strong absorption of blue and ultraviolet light. Melanocytes absorb most of this light. It behaves like blue and ultraviolet filter, which characteristics depend on concentration of melanocytes. Within the epidermal layer there is very little scattering, with the small amount that occurs being forward directed. The result is that all light not absorbed by melanin can be considered to pass into the dermis. By this description, epidermis can be characterised by wavelength dependant coefficients: absorption coefficient of melanin mam(l) and melanin concentration cm [5].

Dermis consist of two sub layers: papillary dermis and reticular dermis. Dermis itself consists of collagen fibres and, in contrast to the epidermis; it contains sensors, receptors, blood vessels and nerve ends. In papillary dermis the collagen fibres are thinner and it behaves as highly backscattering layer. Any incident light is backscattered towards surface. Scattering is grater in red spectrum and going greater to infrared. Because infrared is not absorbed by melanin and blood, this part of spectrum is best for assessing thickness of papillary dermis. Papillary dermis can be described by the absorption coefficients for haemoglobin mah(l), the haemoglobin concentration ch, the scatter coefficient for collagen mspd, and the thickness of collagen layer dpd.

Within the reticular dermis, the large size of collagen fibre bundles causes highly forward-directed scattering. Thus any light, which gets to this layer, is passed deeper into the skin and does not contribute to the spectrum remitted from the skin. Reticular dermis can be described by scatter coefficient msrd, and thickness of layer drd.

Model based feature detection methods

The parameters, describing skin pigmentation, can be found by modelling light transport through layered skin model. To implement this model there can be used Kubelka-Munk theory (Egan and Hilgeman, 1979). It calculates the remitted and transmitted light separately for each layer. And using Beer-Lambert theory of light absorption, there can be calculated concentrations of pigments in each layer, when reflectance coefficients are known [5].

This physical interpretation would let mathematically to define the colour changes because of melanin concentration changes in each layer.

The bigger part of melanoma diagnostic information is hidden in dermis. In more than 90 % skin cancer cases the dermal melanin causes melanoma [6].

The biggest problem is to define the melanin concentration in reticular dermis because almost all light is backscattered from papillary dermis and the rest of light is forward scattered into deeper tissues almost without reflection. To solve this problem there should help Beer–Lambert law combined with Monte-Carlo simulation. Monte-Carlo simulation itself requires much of calculations, but combining with Beer–Lambert theory, this time can be significantly reduced. When there are known characteristics of upper layers by using Monte–Carlo simulation there can be calculated optical properties of reticular dermis with adequate probability.

The main equations used in skin modelling calculations are [6]:

d – layer thickness, k ≈ μa – absorption coefficient, s ≈ μs- scattering coefficient of each layer, Ri and Ti – represents reflected and transmitted light intensities in i-th layer of layered system.

I found on matlab very handy tool which allows easily to trace boundaries of objects in a picture. So I adopted it to skin lesions. This can be used for automatic detection of skin irregularities and used to calculate lesion properties like asymmetry of shape, or border irregularities, who can help in detecting melanoma. There are numerous of investigations done, so I only put few examples on how it looks like. I will give you my source code, so you can try it by your own.

Look at my results:

1) 

And it also finds center of mass;

 2)

and center of mass:

3)

Here is my source code:

%———————————-

clear all
close all
clc
%k parameter can be changed to adjust intensity of image
ei=25;
st=35;
%k=10
k=ei*st;
I = imread(’1.jpg’);
%h=filter matrx
h = ones(ei,st) / k;
I1 = imfilter(I,h,’symmetric’);
figure
subplot(2,2,1),imshow(I), title(’Original image’);
subplot(2,2,2), imshow(I1), title(’Filtered Image’);
IG=rgb2gray(I1);
%Converting to BW
I11 = imadjust(IG,stretchlim(IG),[]);
level = graythresh(I11);
BWJ = im2bw(I11,level);
dim = size(BWJ)
IN=ones(dim(1),dim(2));
BW=xor(BWJ,IN);  %inverting
subplot(2,2,3), imshow(BW), title(’Black and White’);
%Finding of initial point
row = round(dim(1)/2);
col = min(find(BW(row,:)))
%Tracing
boundary = bwtraceboundary(BW,[row, col],’W');
subplot(2,2,4),imshow(I), title(’Traced’);
hold on;
%Display traced boundary
plot(boundary(:,2),boundary(:,1),’g',’LineWidth’,2);
hold off
% figure
% plot(boundary(:,2),boundary(:,1),’black’,'LineWidth’,2);

nn=size(boundary);
KM=zeros(dim(1),dim(2));
 ii=0;
 %Create new matrix with boundary points. there fore we can get rid off
 %other distortions outside boundaries
 while ii<nn(1)
     ii=ii+1;
    KM(boundary(ii,1),boundary(ii,2))=1;
end
 figure
 subplot(2,2,1),plot(boundary(:,2),boundary(:,1),’black’,'LineWidth’,2);
 subplot(2,2,2),imshow(KM)
%Fill inner boundaries where lesion is located
KM2 = imfill(KM,’holes’);
subplot(2,2,3),imshow(KM2)
KM1=xor(KM2,IN);
% subplot(2,2,4),imshow(KM1)
%Geometrical center
IVx=[1:dim(2)];
IVy=[1:dim(1)];
IMx=ones(dim(1),1)*IVx;
IMy=ones(dim(2),1)*IVy;
IMy = imrotate(IMy,-90);
Koordx=IMx.*KM2;
Koordy=IMy.*KM2;
xmean=mean(Koordx,2);
yc=round(sum(xmean.*IMy(:,1))/sum(xmean));
ymean=mean(Koordy);
xc=round(sum(ymean.*IVx)/sum(ymean));
figure
imshow(I)
hold on
plot(boundary(:,2),boundary(:,1),’green’,'LineWidth’,2);
hold on
plot(xc,1:dim(1),’red’,'LineWidth’,2);
plot(1:dim(2),yc,’red’,'LineWidth’,2);
hold off
% ID=im2double(I);
ID1(:,:,1)=im2double(I(:,:,1));
ID1(:,:,2)=im2double(I(:,:,2));
ID1(:,:,3)=im2double(I(:,:,3));
 figure
subplot(2,2,1), imshow(ID1);
subplot(2,2,2), imshow(ID1(:,:,1));
hold on
plot(xc,1:dim(1),’red’,'LineWidth’,2);
plot(1:dim(2),yc,’red’,'LineWidth’,2);
hold off
subplot(2,2,3), imshow(ID1(:,:,2));
subplot(2,2,4), imshow(ID1(:,:,3));

%———————————

To improve diagnostic accuracy the  ABCD rule of lesion screening is widely used based on  asymmetry (A), border (B), color (C), and differential structure (D) measuring.

•         A total dermatoscopic value (TDV) results from the calculation

TDV = A·1,3 + B·0,1 + C·0,5 + D·0,5

•         This score contributes to the differentiation between benign and malignant lesions:

                                           1,00  -  4,75 – benign skin lesion

                                           4,75  -  5,45 – suspicious

                                           More than 5,45 – melanoma

Asymmetry

A – Asymmetry of Shape, Structure and Color. The lesion is divided in four regions and there is symmetry inspected across x or y axis. If asymmetry is only by one axis it gives 1 point and if on both axis there are 2 points calculated. So for shape, structure and color, there can be 6 points maximum.

 

Border and Color

For measuring border irregularities the lesion is divided in 8 regions. If in one particular region color is ends strictly with boundary of lesion there is one point. If lesion color changes smoothly to skin color then there is 0 points. So for B there can be maximum 8 points.

For color there can be as many points as many colors there can be found on lesion. Usually there are 6 main colors, so total 6 points for C.

 

Dermatological Structure D

Thera are 5 main structures noticed in lesions. No structures – smooth color – 1 point, net structure another 1 point, tree like boundaries another 1 point, dots – 1 point and globules 1 point. Total 5 points can be. In some literature there can be diameter of lesion measured for D. If diameter is more than 6 mm it can be stated as suspicious.

 

When looking in Total Dermatological Value formula, there are obvious that the main criteria are Asymmetry. It is usually the main screening option.  It is easy to inspect yourself using this methodology. If there is any concern, it is better to visit your doctor.