Abstract
The multi-feature fuzzy recognition (MFFR) algorithm was presented to infer the vessel structures, in the context of X-Ray Angiograms (XRA) of the coronary artery. In the modeling, a multi-feature metrics (MFM) was firstly established to describe the local configuration; then the membership degree of MFM-based fuzzy subsets was defined, and the fuzzy recognition operator was constructed. The MFFR algorithm can correctly infer four kinds of vessel structures including vascular ends, segments, bifurcations and crossovers. The results are satisfying: on average 91.1% of the testing vessel lengths in medium quality images are automatically delineated as well as their structures being correctly inferred with point-wise.KeywordsMulti-feature fuzzy recognitionvessel structure inferencecoronary artery X-ray angiograms
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
