Fuzzy classification of renal artery lesions from angiograms in the context of angioplasty
Jaulent, M.C.; Cherrak, I.; Salamé, T.; Beyssen, B.; Battaglia, C.; Plouin, P.F.; Degoulet, P.
Studies in Health Technology and Informatics 43 Pt B: 492-496
1997
ISSN/ISBN: 0926-9630 PMID: 10179714 Document Number: 474401
This paper describes a system that automatically classifies renal artery lesions from arteriograms in the context of Percutaneous Transluminal Renal Angioplasty (PTRA). PTRA outcomes have been categorized into type A (success) or type B (failure) according to morphological features. The morphological features are naturally represented by fuzzy sets. The classification process is based on a fuzzy pattern matching procedure between fuzzy prototypes of lesions in the classes A and B and fuzzy quantified assessments of lesions in angiograms. The system provides a meaningful membership degree to the class rather than an abrupt yes or no answer. The system is expected to be a good framework to study the technical results of PTRA according to the anatomy of the lesion.