Design of optical imaging based vascular evolution monitoring system for superficial tumor medication

Visibility of human subcutaneous tissue vascular structure is very important in medicine for vascular surgery operation and fluid administration. Detection of venns with human eye is related with some factors. Visibility of venns can change with tissue thickness, fatty tissue proportion, radius of vessels, age and skin color. In the case of subcutaneous compounds (lipids, water, melanin vb.) absorbs more light than hemoglobin, vessels can not be detected because of inadequate contrast. Hemoglobin has higher absorbing coeeficient in near infrared region so that in this thesis NIR LED arrays are used for illuminating to forearm. When skin has been illuminated by these light sources, vascularity of illuminated region seemed like dark lines in the image because of higher absorbance of haemoglobin. Imaging set-ups have been searched fot the best contrast quality resulted image by performing experiments. To inversitegate how much deepest venns can be imaged, tissue-vessel mimicking phantoms have been prepared. With this set-up it has been observed that venns at 4mm deep can be viewed in image. NIR images have been processed with image enhancement algorithms to get binary images. Radius of vessels have been calculated in binary images. After that, binary images have been processed with thinning algorithm to calculate vascular complexity