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:
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.
There are many skin image capture methodologies developed and used. Here are short review of them: Dermatoscopic photography The deepest layer of skin can be reached – Papillary dermis Resolution – depends on optical system View of skin – Horizontal The main disadvantage is reflections of light from skin surface – stratum cornea. Dermatoscopic oil immersion photography The deepest layer of skin can be reached – Papillary dermis Resolution – depends on optical system View of skin – Horizontal Reflections of light from skin surface are smaller because of oil used between camera optics and skin.
I used simple lamp directed to glossy table surface. One polarizer is in front of lamp and other is in front of lens of digital camera. Both polarizes are perpendicularly oriented to each other. How does this work? There is a theory about an angle on which the incident polarized electromagnetic waves turn reflects from surface with polarization plane turned in 90 degrees. When light going through polarizer towards the surface, the light is polarized in one direction and when it reflects from surface it is turned by 90 degrees and those waves are filtered by another perpendicular polarizer in front of lens.
Let’s make a filter, which filters off the 60Hz frequency from ECG signal. As we know American power supply is 60Hz. This is common noise in biomedical signals, while they are powered from industrial power supply. This type of noise can be defined easily and can be filtered as parameters of noise are known. Here is one example of how to implement FIR filter using mathematical tools, like Matlab. This can be done by using microcontroller, like ARM or even ARM, because the frequencies are up to 1 KHz. Initial conditions: f0=60Hz – pover supply frequency; fs=500Hz – sampling rate; frequencies who define complex zeros: we get w0=0.754; Positions of complex zeros: Zeros and poles in z plane System Function From it we can calculate filter coefficients: And filter coefficients: Â Â Â Â Â Â Also we know that: And here we get filter characteristics: We have band stop filter at 60Hz and…
To evaluate skin pigmentation in different skin layers, there is special light adapter needed in order to take multispectral pictures of skin. As there are different optical properties of skin pigments, 4 different light sources have been chosen. blue λ= 470 nm – highly absorbed by epidermal melanin green λ= 576 nm – hemoglobin peak red λ= 660nm – epidermal dermal boundary IR λ= 865 nm – low absorption, sensitive to scattering to measure papillary dermis thickness. There was lighting source for “Nikon Coolpix E3100” digital camera developed to take multispectral images of skin. Making adapter The drawing of lighting adapter The lighting adapter isn’t very hard to build. You just need to make circular PCB and solder LEDs with protective resistors. The PCB image: