The Evaluation of Fetal Cardiac Remote Screening in the Second Trimester of Pregnancy Using the Spatio-Temporal Image Correlation Method

We prospectively performed remote fetal cardiac screening using the spatio-temporal image correlation (STIC), and examined the usefulness and problems of remote screening. We performed heart screening for all pregnant women at four obstetrics clinics over the three years from 2009 to 2014. The STIC data from 15,404 examinations in normal pregnancies (16–27 weeks, median 25 weeks) were analyzed. Obstetricians and sonographer collected STIC data from four-chamber view images. Eight pediatric cardiologists analyzed the images offline. A normal heart was diagnosed in 14,002 cases (90.9%), an abnormal heart was diagnosed in 457 cases (3.0%), and poor images were obtained in 945 cases (6.1%). 138 cases had congenital heart disease (CHD) after birth, and severe CHD necessitating hospitalization occurred in 36 cases. We were not able to detect CHD by screening in 12 cases. The sensitivity and specificity of STIC in CHD screening was 50% and 99.5%, respectively. The sensitivity and specificity of STIC in screening for severe CHD was 82% and 99.9%, respectively. The STIC method was useful in fetal remote screening for CHD. However, the fact that > 10% of images that could not be analyzed by this method was a problem.