Purpose: This study aimed to evaluate the accuracy of prenatal diagnosis using a spatio-temporal image correlation-based telediagnosis system at a tertiary referral unit in a Japanese prefecture over a 9-year period, and to investigate the validity of delivery hospital selection.
Methods: This study retrospectively analyzed telediagnosis cases from 2009 to 2018 in six geographically remote hospitals. We built a telediagnosis system called the Kyoto Fetal Ultrasonographic Telediagnosis System (K-FUTS). Three-dimensional ultrasonographic images from these referral hospitals were forwarded to the Kyoto Prefectural University of Medicine (KPUM) Hospital through an optical fiber network system for analysis. The primary endpoint was accuracy of prenatal diagnosis. Cases with severe congenital heart disease (CHD) that required immediate postnatal treatment were identified using the K-FUTS. These cases were then examined to determine whether they were delivered at appropriate hospitals.
Results: During the study period, datasets from 182 cases were transferred to KPUM Hospital. Twenty-one datasets were excluded because of data unavailability or poor image quality. CHD was detected in 14.9% of cases (24/161); the accuracy of prenatal diagnosis was 95.0% (153/161). The K-FUTS identified seven severe cases with CHD requiring immediate postnatal surgical or medical treatment. These cases were delivered at KPUM Hospital with a pediatrician present. The remaining 17 cases considered suitable for delivery at the referral hospitals were delivered there, and they did not require intervention immediately after birth.
Conclusion: Our telediagnosis system contributed to the prenatal diagnosis of CHD and management of pregnancies in patients who were geographically remote from a tertiary hospital.
Keywords: Congenital heart disease; Prenatal diagnosis; Spatio-temporal image correlation; Telediagnosis system.