Evaluation Of Fetal Left Ventricular Volume With Spatiotemporal Image Correlation (STIC)

Objective:The purpose of this study was to investigate the feasibility of using novel three-dimensional echocardiographic technique-spatiotemporal image correlation (STIC) to assess left ventricular volume in normal fetuses.Methods:Study Populations and Methods:From October 2009 to February 2010, we randomly enrolled 162 consecutive pregnant women, of whom 134 were recruited into study group (age 27.1±4.8 years, range 21～38 years, gestation age 30.1±4.2 weeks, range 22～38 weeks, fetal heart rate 145±9.1bpm,rang 124～162bpm).Twenty-six fetuses with poor 4D STIC image and 2 twin pregnancies were excluded.4D volumes of the fetal heart were acquired by sonography using a motorized convex transabdominal transducer (RAB 4-8MHz probe, Voluson GE E8,GE Medical Systems) and data were processed by a sonography system capable of performing the STIC technique. First, the 2D settings were adjusted in order to obtain clear visualization of the internal limits of the heart. We aimed to acquire the volume when the fetus was not moving and the apex of the heart was towards the transducer, which was held over the transverse plane of the fetal heart at the level of the standard four-chamber view. Acquisition time ranged from 7.5s to 12.5s and the angle of acquisition ranged between 30°～75°depending on fetal montion and gestational age.The volumes with clear internal limits and no artifacts were then selected for further evaluation.Offline analysis was performed using Voluson 4D View software.The Virtual Organ Computer-aided Analysis (VOCALⅡ),M-mode Teichholz formula and 2D biplane Simpson's method were then used to evaluate end-systolic and end-diastolic volumes of left ventricle.In this study the VOCALⅡwas used to evaluate volumes of left ventricle by obtaining a sequence of 12 sections of the heart, each obtained after a 15°rotation from the previous one. In each plane the contour was traced manually,and at the end, the computer provided the reconstruction of the left ventricle and calculated its volume.Every measurement was done offline after the scan by the same operator for three times.2. In 20 randomly selected cases, LVEDV and LVESV were measured by two independentechocardigraphers in order to compare the measurements and calculate interobserver agreement.In 20 randomly selected cases, LVEDV and LVESV were measured by the same echocardigrapher twice after one month in order to compare the measurements and calculate intraobserver agreement.Statistical analysisThe data were analyzed using the statistical software package SPSS 16.0. A value of P<0.05 was considered statistically significant.Results:1.Result of the reproducibility:For the interobserver LVEDV and LVESV correlation coefficients were 0.988 and 0.887 in 4D STIC VOCAL II rule,0.982 and 0.883 in Simpson's method,0.909 and 0.843 in Teichholz formula. In the Bland-Altman test the bias were 0.190ml and 0.028ml in 4D STIC VOCALⅡrule for LVEDV and LVESV, 0.222ml and 0.110ml in Simpson's method,0.691ml and 0.361ml in Teichholz formula. For the intraobserver LVEDV and LVESV correlation coefficients were 0.998 and 0.977 in 4D STIC VOCALⅡrule,0.996 and 0.962 in Simpson's method,0.937 and 0.928 in Teichholz formula. In the Bland-Altman test the bias were 0.076ml and 0.053ml in 4D STIC VOCAL II rule for LVEDV and LVESV,0.102ml and 0.076ml in Simpson's method,0.430ml and 0.263ml in Teichholz formula.2.The highest concordance correlation (r=0.968 for LVEDV,r=0.956 for LVESV; P=0.000) was noted between Simpson's and 4D STIC VOCAL II values with a small bias (0.282ml for LVEDV,0.117ml for LVESV). There were poor concordance correlation between Simpson's method and M-mode Teichholz formula,0.757 for LVEDV and 0.552 for LVESV with bias 0.809ml for LVEDV,0.419ml for LVESV,so did to 4D STIC VOCAL II and M-mode Teichholz formula,0.742 for LVEDV and 0.491 for LVESV with bias 0.866ml for LVEDV,0.463ml for LVESV.3.The Bland-Altman analysis showed that 4D STIC VOCAL II had the best agreement than M-mode Teichholz formula and Simpson's method. The result showed that there were significant difference between M-mode Teichholz formula and Simpson's method in LVEDV and LVESV (P<0.05).There were significant difference between M-mode Teichholz formula and 4D STIC VOCALⅡ(P<0.05), too. There were no significant differences between 4D STIC VOCAL II and Simpson's method in LVEDV and LV ESV.Conclusion:There is a good agreement between left ventricular volumes measured either by 4D STIC VOCAL II or by 2D biplane Simpson's method.4D STIC is a novel technique that can be more accurate than conventional methods in quantification of fetal left ventricular volumes. All the volume data acquisition is achieved by postprocessing facilities.STIC sweep decreases the exposure upon fetal heart obviously.4D STIC promises to become a new method for left ventricular.