Diagnosing renal artery lesions with a fuzzy logic model
Marie‐Christine Jaulent, Patrice Degoulet · IEEE Engineering in Medicine and Biology Magazine · 1994
The authors present a fuzzy classification procedure based on a well known fuzzy pattern matching technique (H. Dubois et al., Fuzzy Sets and Systems, vol. 28, p. 313-33, 1988). This procedure has been developed in order to diagnose the visual type of a lesion from 2D and static renal artery angiograms. In the implemented model, each observed lesion no longer belongs entirely to only one class (a specific visual type) but rather to all classes, with differing membership values. Here, the authors first place this concept in the general framework of computer systems that take imprecision and uncertainty into account in medical reasoning. They then present the method that they have developed in the specific context of diagnosing renal artery lesions and they describe the classification procedure itself, which is derived from Dubois et al.>