| 王学东,刘爱连,张钦和,王家正,陈丽华,王楠,宋清伟,王易世.压缩感知技术对3D mDIXON Quant定量分析肝脏脂肪的影响[J].中国医学影像技术,2020,36(11):1662~1666 |
| 压缩感知技术对3D mDIXON Quant定量分析肝脏脂肪的影响 |
| Impact of compressed sensing technology on 3D mDIXON Quant liver fat quantification |
| 投稿时间：2019-09-01 修订日期：2020-04-23 |
| DOI：10.13929/j.issn.1003-3289.2020.11.016 |
| 中文关键词: 脂肪肝 脂肪组织 磁共振成像 压缩感知 |
| 英文关键词:fatty liver adipose tissue magnetic resonance imaging compressed sensing |
| 基金项目:首都科技领军人才培养工程（Z181100006318003）。 |
|
| 摘要点击次数: 2437 |
| 全文下载次数: 789 |
| 中文摘要: |
| 目的 探讨压缩感知（CS）技术不同加速因子对3D mDIXON Quant定量分析肝脏脂肪的影响。方法 对20名成人志愿者行上腹部MRI，扫描序列包括传统SENSE-3D mDIXON Quant（SENSE组）和不同加速因子（2、4、5、6）CS-3D mDIXON Quant（CS2、CS4、CS5、CS6组），记录各组的扫描时间。经分析获得脂肪分数图，由2名医师分别于肝门水平肝左外叶、左内叶、右叶前段和右叶后段测量脂肪分数。采用组内相关系数（ICC）分析2名医师测量结果的一致性。比较不同CS组与SENSE组脂肪分数的差异；采用Bland-Altman法分析不同CS组间脂肪分数的一致性。结果 2名医师测量各组脂肪分数的一致性较好（ICC均≥0.98，P均<0.01）。SENSE组扫描时间为13.01 s，CS2、CS4、CS5及CS6组扫描时间分别为15.02 s、7.69 s、6.18 s及5.10 s，其脂肪分数与SENSE组间差异均无统计学意义（Z=－0.07、－0.74、－0.34、－0.14，P均>0.05）。Bland-Altman图显示，不同CS组之间脂肪分数一致性均较好。结论 CS技术结合mDIXON Quant序列可在不影响肝脏脂肪定量分析结果的前提下显著缩短扫描时间。 |
| 英文摘要: |
| Objective To explore the impact of compressed sensing technology (CS) on 3D mDIXON Quant liver fat quantification . Methods Abdominal MR images were performed in 20 adult volunteers. The scanning sequences included traditional SENSE-3D mDIXON Quant (SENSE group) and different accelerators (2, 4, 5, 6) CS-3D mDIXON Quant (CS2, CS4, CS5, CS6 group), and the scanning time of each group was recorded and analyzed, and the fat score map was obtained. The fat fraction of liver was measured on the left outer lobe, left inner lobe, anterior and posterior segments of the right lobe at the hepatic hilar level by two imaging physicians. Intra-class correlation coefficients (ICC) test was used to analyze the consistency of the measurement results between two physicians. The fat fractions of different CS groups were compared with that of SENSE group, respectively. Bland-Altman method was used to analyze the consistency of fat fraction between each two CS groups. Results The consistency of fat fraction measured by 2 physicians was good (all ICC≥0.98, all P<0.01). The scanning time of SENSE group was 13.01 s, of CS2, CS4, CS5 and CS6 group was 15.02 s, 7.69 s, 6.18 s and 5.10 s, respectively, all were not statistically different with that of SENSE group (Z=－0.07, －0.74, －0.34 and －0.14, all P>0.05). Bland-Altman plot showed that the consistency of fat fraction between each two CS groups was good. Conclusion CS technology combined with mDIXON Quant could significantly shorten scanning time without affecting liver fat quantification. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
