Radiologist model for cardiac rest period determination based on fuzzy rule

Image data acquisition for the coronary arteries is generally implemented during the diastole rest period, in order to suppress blurring due to cardiac movement. The purpose of this study is to improve the semi-automated application to determine the cardiac rest period based on fuzzy logic. The cardiac rest period from 25 subjects were determined based on their normalized cross-correlation of consecutive frame images as well as normalized frame number as the measured variables. The fuzzy set and membership are generated based on the measured variables from the radiologist's visual assessment. That visual assessment is also regarded as a gold standard for verification. The distance difference between the proposed method and visual assessment was analyzed. The fuzzy logic approach for cardiac rest period determination has no significant difference compared to the visual assessment (p>0.05) in terms of start frame and end frame. The algorithm could be extended easily in case of there are some necessary variables should be added to accommodate rest period definition from different radiologist.