| 摘要: |
| 目的采用两次激发（2-shot）压缩感知（CS）心脏电影成像（cine）技术进行心脏磁共振（CMR）检查，以快速采集能够快速准确评估心功能、更接近标准cine序列的高质量图像。方法连续纳入2018年7月至9月于中国医学科学院、北京协和医学院北京协和医院放射科进行CMR检查的患者。所有纳入患者均完成标准cine序列、单次激发（ss）CS cine序列、2-shot CS cine图像采集。对不同序列采集图像进行图像质量评分和心功能定量分析。结果 22例患者完成3种cine序列成像。标准cine序列、ss CS cine序列、2-shot CS cine序列的平均采集时间分别为（102.50±18.55）、（20.50±3.71）、（30.75±5.57）s，采集图像的总体质量评分中位数分别为4、3、4分，标准cine序列与ss CS cine序列图像的总体质量评分差异有统计学意义（Z=－2.828，P=0.005），而标准cine序列与2-shot CS cine序列的评分差异无统计学意义（Z=－1.000，P=0.317）。在心功能定量研究方面，标准cine序列与ss CS cine序列之间除左心室心肌质量[分别为100.25（47.30～136.02）g和101.25（48.30～135.30）g]差异有统计学意义（Z=－2.484，P=0.013）外，其余指标差异均无统计学意义（P均>0.05）。而标准cine序列与2-shot CS cine序列之间各项心功能参数的差异均无统计学意义（P均>0.05）。Spearman相关分析显示，标准cine序列与ss CS cine序列之间、标准cine序列与2-shot CS cine序列之间，各项心功能参数均有良好的相关性（P均<0.01）。结论 2-shot CS cine序列与ss CS cine序列的采集时间均较标准cine序列明显缩短，但2-shot CS cine序列能够获得更接近标准cine序列质量的图像，心功能参数评估也更准确。 |
| 关键词: 心脏病 多次激发 单次激发 压缩感知 电影磁共振成像 |
| DOI：10.16781/j.0258-879x.2019.03.0257 |
| 投稿时间：2018-10-29修订日期：2019-01-15 |
| 基金项目:国家自然科学基金（81471725），科技部“十三五”国家重点研发计划（2016YFC1300402），2017年度北京市科技新星计划交叉学科合作课题（Z1711100001117136）. |
|
| Two-shot compressed sensing techniques accelerate cardiac cine sequence acquisition and quantitative evaluation of the diagnostic efficacy |
| WANG Jian1,LI Xiao1,LIN Lu1,CAO Jian1,DAI Jing-wen1,AN Jing2,JIN Zheng-yu1,WANG Yi-ning1* |
(1. Department of Radiology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100730, China;
2. Siemens Healthineers, Shenzhen 518057, Guangdong, China
*Corresponding author) |
| Abstract: |
| Objective To perform cardiac magnetic resonance (CMR) examination with 2-shot compressed sensing (CS) cardiac cine imaging (cine) technique, so as to obtain high-quality images that can quickly and accurately assess cardiac function and are closer to the standard cine sequence. Methods The patients, who underwent CMR examination in Department of Radiology of Peking Union Medical College Hospital of Chinese Academy of Medical Sciences & Peking Union Medical College from Jul. to Sep. 2018, were continuously enrolled. All enrolled patients received the standard cine sequence, single-shot (ss) CS cine sequence and 2-shot CS cine sequence image acquisition. Quantitative analyses of image quality and cardiac function were performed on each cine image acquired from different sequences. Results Twenty-two patients completed the three types of cine sequences. The average acquisition time of standard cine sequence, ss CS cine sequence and 2-shot CS cine sequence was (102.50±18.55), (20.50±3.71) and (30.75±5.57) s, respectively. The median total image quality scores of the standard, ss CS and 2-shot CS cine sequences were 4, 3 and 4, respectively. There was significant difference in the total quality scores between the standard cine sequence and the ss CS cine sequence (Z=-2.828, P=0.005), while there was no significant difference between the standard cine sequence and the 2-shot CS cine sequence (Z=-1.000, P=0.317). In the quantitative study of cardiac function parameters, left ventricular myocardial mass was significantly different between the standard cine sequence and the ss CS cine sequence (100.25[47.30-136.02] g vs 101.25[48.30-135.30] g, Z=-2.484, P=0.013), but there were no significant differences in the other parameters. There were no significant differences in the cardiac function parameters between the standard cine sequence and the 2-shot cine sequence. Spearman correlation analysis showed that there were good correlations in the cardiac function parameters between the standard cine and the ss CS cine, and the standard cine and the 2-shot CS cine (all P<0.01). Conclusion The acquisition time of 2-shot CS cine and ss CS cine sequences is significantly shorter than that of standard cine sequence. However, 2-shot CS cine sequence can obtain images closer to the images obatined by standard cine sequence, which can be used to accurately evaluate the cardiac function. |
| Key words: heart diseases multi-shot single-shot compressed sensing cine magnetic resonance imaging |
