Application Research On The Algorithms Of NIRS Breast Image Enhancement And Tissue Oxygenation Detection

In recent years, breast cancer has occupied the first position of thefemale cancer incidence, early diagnosis and treatment of breastcancer is crucial. Using Near infrared spectroscopy scans to check is theprimary means of breast disease, with the advantages of convenience,low prices, the human body without injury and others, which is widelyused in breast disease screening and physical examination. However,because of various factors influence, the images obtained by infraredbreast scan equipment can not meet the diagnostic needs.In order to solve the low contrast images for lesion morphology andvascular information fuzzy problem of the near-infrared breast images,this paper presents a stochastic resonance algorithm based on PSNRthreshold. The algorithm is the different intensity of the Gaussian randomnoise added to the low-contrast images.When the maximum PSNR of thenoisy image and its corresponding optimal MSE is found, the stochasticresonance phenomenon of the image happens, and there has thehighest image contrast. Stochastic resonance converts the image noiseenergy into useful signal energy, which provides a new method forenhancing the near-infrared breast image low-contrast.After the Near-infrared breast image enhancing, it can improve thestructural information of the breast, which provides a favorable basis forthe diagnosis, but in early breast cancer lesions and atypical images ofbreast cancer, this information is not very prominent, so that leads to falsenegative diagnosis. In order to improve the detection of early lesionspositive rate, this thesis improves the oxygenation detection algorithm,and it can calculate the absolute amount of the breast blood oxygencontent. The thesis also designs a dual-wavelength oxygenationdetection system, and uses the breast tissue oxygenation detectionalgorithm to obtain the two-dimensional distribution maps of breast bloodcontent and oxygen saturation, which can more accurately reflect thefunctional lesions of the mammary gland and has very useful value for thebreast cancer diagnosis. Using the detector of blood concentration andoxygen saturation system for180patients’ cases with mammary diseases,and the diagnosis coincidence rate with the establishing pathological diagnosis, and surgical cases is94.7%. The NIRS testing on the bloodconcentration and oxygen saturation technology have sensitivity andclinical result for tumors of breast.