Cardiac screening by STIC: can sonologists performing the 20-week anomaly scan pick up outflow tract abnormalities by scrolling the A-plane of STIC volumes?

Abstract

Objective: To assess whether medically qualified sonologists with low-to-intermediate scanning experience are able to detect major abnormalities of the outflow tracts by reviewing the A-plane of cardiac volume datasets acquired with spatiotemporal image correlation (STIC).

Methods: Fourteen sonologists of low-to-intermediate scanning experience were recruited among residents and colleagues involved in the screening ultrasound clinic at our referral center. Basic criteria for selection were: ability to perform the 20-week anomaly scan and to assess the four-chamber view, inability to perform extended cardiac screening (outflows); willingness to participate in the study. These sonologists attended a 2-hour lesson on: a) how the outflow tract views can be abnormal, and b) how to use a laptop and the dedicated software to review cardiac volumes in the A-plane only. After this briefing, each of them, independently, reviewed 26 preselected volumes at a workstation (from 16 normal fetuses and 10 with outflow tract abnormalities), without knowing how many of them were normal. After reviewing each volume, the sonologist was asked to define the outflow tract views as normal or abnormal and, if willing, to hypothesize the anomaly. The sequence of cases was changed for each participant. The time allotted for review of the volumes was 1 hour (about 2 min per case).

Results: Of the 364 diagnoses from review of the volumes, 116 (31.9%) were true positives, 195 (53.6%) were true negatives, 29 (8.0%) were false positives and 24 (6.6%) were false negatives. The sensitivity, specificity and positive and negative predictive values were 83%, 87%, 80% and 89%, respectively. Individual diagnostic accuracy ranged from 66 to 100% (median, 85.5%) and individual detection rate from 50 to 100% (median, 85%). The detection rate per single congenital heart disease ranged from 50% (for TGA with intact ventricular septum) to 100% (for DORV, DORV with pulmonary atresia and TGA with ventricular septal defect). There was no correlation between detection rate and alignment of the four-chamber view with the ultrasound beam (apical vs. transverse).

Conclusions: In this preliminary study, we have demonstrated that sonologists with low-to-intermediate experience of anomaly ultrasound screening in the second trimester and no experience of insonating the outflow tracts were able to identify outflow tract abnormalities by reviewing the A-plane of cardiac volume datasets, after detailed briefing.