医学
右冠状动脉
树(集合论)
放射科
冠状动脉疾病
心脏病学
冠状动脉造影
数学
数学分析
心肌梗塞
作者
Sebastian Reinartz,Markus Winkler,Sascha B. Diefenbach,Thomas Allmendinger,Tobias Penzkofer,Christiane K. Kuhl,Andreas H. Mahnken
标识
DOI:10.1016/j.acra.2016.09.014
摘要
Particularly for patients with heart arrhythmias, conventional BestSystole (BS) and BestDiastole (BD) reconstruction techniques in computed tomography (CT) frequently show artifacts that hinder the readability of the coronary tree. To address this problem, this paper presents an alternative reconstruction method that combines the technique "reconstructions with identical filling" (RIF) with motion mapping: This new technique is called "RIF in motion mapping" (RIMM). This study compares the diagnostic quality of images generated with RIMM to that of the other reconstruction techniques.Having shown major artifacts in standard reconstructions, the CT datasets of 23 patients with suspected coronary artery disease or prior to transcatheter aortic valve replacement were selected manually. Each dataset was evaluated with four reconstruction techniques: BS, BD, RIF, and RIMM. Two radiologists, blinded to the applied reconstruction type, then evaluated the entire coronary tree of each sample using the 15-segment American Heart Association model and the six-grade Likert scale.Of the 345 analyzed coronary segments, the RIMM technique showed a significant number of images with reliable diagnostic quality (n = 228, 66%) as compared to RIF (P = 0.002) and BS/BD reconstructions (P < 0.001). Per coronary segment, vessel, and patient, the RIMM technique scored significantly better than the conventional BS/BD reconstructions (P = 0.003) and better than the RIF reconstructions with regard to the right coronary artery (P = 0.041).This new technique works: Using RIMM on the worst CT images substantially erased many of these artifacts, thereby enabling the radiologists to clearly visualize these segments. As RIMM considerably eliminates artifacts, this new CT reconstruction technique can help make a fast reliable evaluation of a patient's coronary tree. Thus, this enhanced visualization of cardiac images by RIMM avoids the need for further invasive diagnostic procedures.
科研通智能强力驱动
Strongly Powered by AbleSci AI