The reproducibility of left ventricular strain measured by CT and its agreement with speckle tracking echocardiography / 中华放射学杂志

ABSTRACT

Objective:To investigate the reproducibility of left ventricular strain assessed by CT feature tracking(CT-FT) and its correlation and agreement with speckle tracking echocardiography (STE).Methods:Thirty outpatients with suspected coronary heart disease who underwent whole cardiac cycle coronary CTA and transthoracic echocardiography within one week were prospectively enrolled in November 2019. Left ventricular volume and strain parameters were measured by CT-FT and STE, including left ventricular end-diastolic volume (EDV), end-systolic volume (ESV), ejection fraction (EF), global longitudinal strain(GLS), global radial strain (GRS),and global circumferential strain(GCS). GLS included endocardial global longitudinal strain (EndoGLS) and myocardial global longitudinal strain (MyoGLS), GCS included endocardial global circumferential strain (EndoGCS) and myocardial global circumferential strain (MyoGCS). ICC was used to evaluate intra-and inter-observer differences in strain measured by CT-FT. The differences of measurements between CT-FT and STE were compared by paired-samples t test. Pearson correlation coefficient was used to analyze the correlation between CT-FT and STE measurements. Agreement between measurements of two modalities was assessed by Bland-Altman analysis. Results:There was a good consistency in EDV, ESV, EF, EndoGLS, MyoGLS, GRS, EndoGCS and MyoGCS measured by CT-FT between intra-and inter-observer (ICC was 0.775-0.964, P<0.001). There was no significant difference in EF measured by CT-FT and STE [(60.27±8.70) % and (61.22±5.64) %, P=0.443]. EndoGLS, MyoGLS, GRS and MyoGCS measured by CT-FT were (-20.47±4.01)%, (-18.06±3.75)%, (73.90±20.58) % and (-18.48±3.65)%, respectively, while the strain measured by STE were (-18.97±3.33)%, (-16.49±2.60)%, (18.56±3.06)% and (-20.26±4.45)%, respectively. The differences were statistically significant between CT-FT and STE ( t=-2.367, -2.945, 12.161, 2.459, P<0.05). The EndoGCS measured by CT-FT and STE were (-27.78±6.66)% and (-29.18±7.24)%, respectively, with no statistical significance ( P=0.223).The correlation coefficients of EndoGLS, MyoGLS, GRS, EndoGCS and MyoGCS measured by CT-FT and STE were 0.566, 0.629, 0.509, 0.606 and 0.539, respectively ( P<0.05). The average difference of EndoGLS, MyoGLS, GRS, EndoGCS and MyoGCS measured by CT-FT and STE was -1.5%, -1.6%, 55.3%, 1.4% and 1.8%, respectively, with 95% limits of agreement was -8.3%-5.3%, -7.3%-4.2%, 18.1%-92.5%, -10.7%-13.6% and -6.0%-9.5%, respectively. Conclusions:The left ventricular global strain evaluated by CT-FT was feasible, and the agreement of global strain between CT-FT and STE was good but not interchangeable. CT can be an alternative method for "one-stop" evaluation of cardiac anatomy and function in patients with poor echogenic windows and contraindications for MRI.