首页|第二代冠状动脉追踪冻结技术结合深度学习图像重建算法在较高心率冠状动脉成像中的应用

第二代冠状动脉追踪冻结技术结合深度学习图像重建算法在较高心率冠状动脉成像中的应用

扫码查看
目的 探索第二代冠状动脉追踪冻结(SSF2)技术,结合深度学习图像重建算法在较高心率患者冠状动脉CT血管成像(CCTA)中的应用价值.方法 回顾性纳入心率≥70 次/min行CCTA检查的患者 62 例,CT扫描管电压为 100 kV.采用双盲法对冠状动脉的图像质量进行主观评价,分为常规图像(A组)、SSF1 图像(B组)和SSF2 图像(C组).结果 三组图像质量评分和优秀率比较显示RCA、PDA、LAD、D、LCX、OM、PLB、RI差异均有统计学意义(P<0.05).冠状动脉的评分为(2.85±0.72)(A组)、(3.38±0.61)(B组)、(3.86±0.34)(C组).冠状动脉的优秀率为 15.19%(A组)、44.39%(B组)、86.42%(C组).可诊断率比较显示RCA、PDA、D、LCX、OM、PLB、RI差异有统计学意义(P<0.05),而LAD可诊断率差异无统计学意义(χ2=4.01,P=0.135).结论 采用SSF2 技术结合深度学习图像重建算法可以进一步提高较高心率患者CCTA图像质量.
Application of the second generation of snapshot freeze combined with deep learning im-age reconstruction algorithm in coronary CT angiography in patients with high heart rate
Objective To explore the application value of the second generation of snapshot freeze(SSF2)combined with deep learning image reconstruction algorithm in coronary CT angiography(CCTA)in patients with high heart rate.Methods A total of 62 patients with heart rate≥70 times/min who underwent CCTA examination were retrospectively included,and the tube voltage was 100 kV.The image quality of coronary artery was evaluated subjectively by double-blind method,which was divided into conventional images(group A),SSF1 images(group B),and SSF2 images(group C).Results The comparison of image quality scores and ex-cellent rates among the three groups showed that RCA,PDA,LAD,D,LCX,OM,PLB and RI were significantly different(P<0.05).Coronary artery scores were(2.85±0.72)(group A),(3.38±0.61)(group B),and(3.86±0.34)(group C).The excel-lence rate of coronary artery was 15.19%(group A),44.39%(group B),86.42%(group C)respectively.The comparison of diag-nosable rates showed that RCA,PDA,D,LCX,OM,PLB and RI were statistically significant(P<0.05),and LAD was not statisti-cally significant(χ2=4.01,P=0.135).Conclusions SSF2 technology combined with deep learning image reconstruction algorithm can further improve CCTA image quality in patients with high heart rate.

snapshot freezedeep learning image reconstructioncoronary CT angiographyimage quality

张秀芳、罗春材

展开 >

100853 北京,解放军总医院第二医学中心放射诊断科

100853 北京,解放军总医院第一医学中心放射诊断科

冠状动脉追踪冻结技术 深度学习图像重建 冠状动脉CT血管成像 图像质量

2024

武警医学
中国人民武装警察部队后勤部卫生部

武警医学

CSTPCD
影响因子:0.747
ISSN:1004-3594
年,卷(期):2024.35(10)