Color-coded Vessel Imaging Based On Adaptive Multi-scale Filtering

This paper researches into the enhancement of digital subtraction angiography sequenceand then the application of color-coded imaging technology to each frame in the sequenceafter preprocessing. Thus the blood flow information that is stored in the whole black andwhite digital subtraction angiography sequence can be displayed in one single color image.Through the above improvement, doctors are able to interpret accurately angiographic resultsand make diagnosis much more quickly and efficiently. This research is conducted as follows:Considering the low gray level of vascular structure in processing vessel enhancementby using the classical multi-scale filtering algorithm, this paper designed an adaptivemulti-scale filter for vessel enhancement. On the basis of the frames implemented withcontourlet transform, the de-noised enhanced vessel images are obtained by applying theadaptive multi-scale filtering process. This paper improves the adaptive initial parameters ofthe algorithm so as to conclude a clearer digital subtraction angiography sequence of a set ofenhanced vessel structure images.In clinical applications, it is common practice for doctors to analyze black and whitedigital subtraction angiography sequence to diagnose a disease, during which diagnosis arefull of trifles and with low accuracy; what’s worse, the diagnosis can be easily affected bydoctors’ experience. This paper proposes an improved color-coded vessel imaging technologyto solve the issues. By applying color-coded vessel imaging technology in the enhancedsequence images, the blood flow information in the whole monochrome digital subtractionangiography frames can be displayed in one single color image in both static and dynamicway. This improvement eases doctors’ burdens of reading sequence images and the make thediagnosis more efficiency; at the same time, it reduces the effect of doctors’ subjectivefeelings on the diagnosis results and enhances the accuracy of disease recognition.