摘要
目的评估人工智能(AI)触发技术在冠状动脉CT血管成像(CCTA)中对血管峰值增强时间预测的准确性,并将其与传统的团注追踪技术进行比较。方法在这项前瞻性研究中,将204名连续患者连续分为两组进行CCTA扫描:以固定的5秒触发后延迟(PTD)为特点的传统团注追踪技术组(A组);以患者特异性PTD为特点的AI触发技术组(B组)。所有扫描参数(管电压100kV结合CARE Dose 4D的智能管电流调节技术)和对比剂注射方案(320mgI/mL;0.7mL/kg)在各组之间保持一致。一位医师评估了客观图像质量,而两位医师使用Likert 5级评分法评估主观图像质量。通过双样本t检验比较各组之间的客观图像质量,同时用卡方检验比较主观评分。结果两组均有102名患者(A组平均年龄50±10岁;B组平均年龄51±9岁),具有可比较的基线特征。两位医师都对B组的主观图像质量进行了评分(P<0.001)。B组所有冠状动脉血管平均血管增强显着增高(P均<0.05)。结论与传统团注追踪技术相比,AI触发技术可以实现可靠的扫描时间,优化血管强化峰值,获得更好的CCTA图像质量。
Objective To validate the peak vessel enhancement timing of artificial intelligence(AI)trigger technique in Coronary CT angiography(CCTA)and compare its subjective and objective image quality against the traditional bolus tracking technique.Methods In this prospective study,204 patients were serially divided into two groups to perform CCTA scans:traditional bolus tracking technique featuring a fix post-trigger delay(PTD)(Group A)or AI trigger technique featuring a patient-specific PTD(Group B).All CT(tube voltage 100kV combined with intelligent tube current adjustment technology of CARE Dose 4D)and contrast media protocol parameters(320 mgI/mL;0.7mL/kg)were kept identical between groups.One reader evaluated objective image quality,while two readers rated subjective image quality in 5-point Likert scale.Objective image quality was compared between groups via twosample t-test,while the subjective ratings were compared with chi-square analysis.Results There were 102 patients in both groups(mean age 50±10 years in group A;mean age 51±9 years in group B).The baseline characteristics were equivalent between groups(all P>0.05).Both readers rated better subjective image quality for the patient-specific PTD group(P<0.001).The mean vessel enhancement was significantly higher in group B in all coronary vessels(all P<0.05).Conclusion Compared with fixed PTD,patient-specific PTD could achieve reliable scan timing,optimize vessel opacification and obtain better image quality in CCTA.
作者
刘杰
原典
张怡存
高剑波
LIU Jie;YUAN Dian;ZHANG Yi-cun;GAO Jian-bo(Department of Radiology,The First Affiliated Hospital of Zhengzhou University,Zhengzhou 450000,Zhengzhou Province,China)
出处
《中国CT和MRI杂志》
2024年第7期83-86,共4页
Chinese Journal of CT and MRI
基金
河南省卫生健康委员会科技攻关项目(212102310142)。
关键词
人工智能
冠状血管
体层摄影术
X线计算机
Artificial Intelligence
Coronary Vessels
Tomography,X-ray Computed