Near Infrared Vein Image Feature Enhancement And Visualization Algorithm

The fast and precise location of the vein is a common requirement in medicalcare. The worldwide patient of about several billions need intravenous access therapyevery day. Traditional method of venous access suffering the impact of medical care’sexperience is difficult to realize, let along the accurate and reliable detection of thedeep vein. Repeated intravenous puncture caused by venous access difficult hasbrought about the difficult in medical treatment and patient’s suffering and even leadto medical dispute. Presently, the near infrared imaging is used to capture the veinimage and the image enhancement algorithm is applied to improve the image qualityand the uneven characteristic. The purpose of this study is to develop a vein featureenhancement and visualization algorithm as an intervention to improve the imagingcharacteristics of venous access auxiliary device.The main research were as follows:1. The establishment of experimental system. Appropriate band was selectedaccording to the spectral characteristics of hemoglobin in human venous blood andthe vein image acquisition system was established.2. The development of imaging algorithm. An image enhancement algorithm wasdesigned and developed for this study, considering the advantage of histogramequalization and the vein feature enhancement algorithm based on the Hessian matrix.Firstly, the contrast limit and the adaptive method was used to improve result ofhistogram equalization. Then the algorithm efficiency was improved based on thebilinear interpolation, so as to promote image contrast rapidly. Finally, the vein featureenhancement algorithm based on multi-scale of Hessian matrix was used to extract thevein information. The experimental results showed that the algorithm can be used forlow contrast image enhancement, the effect of enhancement is clear.3. The feature extraction based on imaging algorithm. A vein featurevisualization algorithm including the display of intravenous width information and theextraction the velocity of blood filling was designed and developed. The veindirection was extracted and the vein width was calculated based on the analyzing ofgradient characteristic on the edge of the blood vessel. Then the dynamic informationof the image was captured by frame-to-frame differences. The distance of blood flowwas extracted through the duty ratio threshold method, thus the venous blood flowvelocity and the veins recovery time per unit length was calculated. The research results showed that the algorithm performed well in the assessment of venous featurewith higher accuracy.The algorithm researched in this study, creates a clear instant picture of the vienand can apply to the vein imaging mode, the width visualization mode and the bloodflow information detection mode. The clinical application in health care shows thatthe algorithm is beneficial to increase the success rate of venous access and beneficialto patients, reducing the difficulty of staff work.