Feasibility of left ventricular longitudinal strain measurement by speckle-tracking echocardiography in infants / 中华超声影像学杂志

ABSTRACT

Objective:To explore the clinical feasibility and application method of speckle-tracking echocardiography (STE) in assessing left ventricular longitudinal strain in infants.Methods:One hundred and ten infants within one week of birth were randomly selected in the Third Affiliated Hospital of Guangzhou Medical University from June 2019 to February 2020, and the basic data were collected. STE was performed by two physicians with more than 5 years′ experience.Images of the same infant were analyzed by physicians with 5 years′ experience and 2 years′ experience, respectively, and the difference and correlation of the results between the two doctors were analyzed. The images of the same infant were analyzed twice by the doctor with 5 years′ experience, and the differences and correlations between the results were analyzed. The infants were divided into different groups according to weight to compare the correlations between inter-physicians and intra-physician. STE analysis indicators included global left ventricular longitudinal peak strain (Glps-avg), four-chamber left ventricular longitudinal strain (Glps-A4C), three-chamber left ventricular longitudinal strain (Glps-A3C), two-chamber left ventricular longitudinal strain(Glps-A2C), left ventricular anterior wall longitudinal strain (Glps-a), anterior interventricular septum longitudinal strain(Glps-ais), posterior interventricular septum longitudinal strain (Glps-pis), inferior wall of left ventricle longitudinal strain (Glps-i), posterior wall of left ventricle longitudinal strain (Glps-p), lateral wall of left ventricle longitudinal strain (Glps-l), left ventricular basal section longitudinal strain (Glps-bs), left ventricular middle section longitudinal strain (Glps-ms) and left ventricular apex section longitudinal strain (Glps-as), a total of 13 indicators. The differences were analyzed by paired t test, and the correlations were determined by intra-group correlation coefficient (ICC). Results:According to the inclusion and exclusion criteria, 95 patients were included. For the 13 STE indicators, inter-physicians comparison the differences between the two physicians were statistically significant (all P<0.05) except for Glps-pis, and the ICC were 0.38-0.72 (all P<0.01). Intra-physician comparison the differences of these indicators between two measurements were not statistically significant (all P>0.05) except for Glps-ais and Glps-I, ICC were 0.31-0.76 (all P<0.05). Among them, inter-physicians and Intra-physician ICC of Glps-avg, Glps-bs, Glps-ms were 0.68/0.75, 0.72/0.66 and 0.65/0.76 respectively. The group comparison by infants′ weight showed that In very low weight infants group, the ICC of inter-physicians and intra-physician ranged 0.82-0.93(all P<0.05) and 0.80-0.95(all P<0.05). In low weight infants group, the ICC of inter-physicians and intra-physician ranged 0.65-0.94 (all P<0.05) and 0.69-0.93 (all P<0.01). In the normal weight infants group, ICC of inter-physicians ranged 0.06-0.68, with statistical significance except for Glps-A3C (all P<0.05); ICC of intra-physician ranged 0.36-0.59 (all P<0.05). In overweight infants group, there was no statistical significance in ICC of inter-physicians (all P>0.05), the ICC of intra-physician, Glps-bs and Glps-ms groups were 0.63 and 0.77 (all P<0.05), with no statistical significance in other indicators (all P>0.05). Conclusions:Left ventricular longitudinal strain by STE in infants has better repeatability, and the consistency of intra-physician is higher than inter-physicians, among which the Glps-avg consistency is the best, followed by Glps-bs and Glps-ms. STE is more reproducible in low and very low weight infants than in normal and overweight infants.