Exploration of intelligent application in quality control of fetal ultrasound examination in early pregnancy in Liaoning province
10.3760/cma.j.cn131148-20250228-00109
- VernacularTitle:辽宁省早孕期胎儿超声检查质量控制的智能化应用探索
- Author:
Zimeng ZHANG
1
;
Fujiao HE
;
Sihong WANG
;
Ying HUANG
;
Lizhu CHEN
Author Information
1. 中国医科大学附属盛京医院超声科，沈阳 110004
- Publication Type:Journal Article
- Keywords:
Ultrasonography;
Fetus;
Artificial intelligence;
Quality control;
Early pregnancy
- From:
Chinese Journal of Ultrasonography
2025;34(7):571-578
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To evaluate the application efficiency of the artificial prenatal sonography quality assurance system（PSAIS）in the quality control of fetal image quality during the first trimester（11-14 weeks），and to assist the quality control of fetal ultrasound examinations in the first trimester in Liaoning Province using this system.Methods:One hundred fetuses（2 757 ultrasound images）from early pregnancy ultrasound screenings at the Shengjing Hospital Affiliated to China Medical University in November 2022 were retrospectively randomly selected as the test set. The performance of PSAIS were evaluated by 3 obstetric ultrasound experts. First，the experts' sectional classification was used as the gold standard，the accuracy of PSAIS in classifying standard sections was assessed；the consistency between PSAIS and expert ratings was evaluated using the intraclass correlation coefficient（ICC）. The difference of time between PSAIS and expert quality control was analyzed. Finally，PSAIS was used to evaluate the quality of early pregnancy fetal ultrasound examinations at various levels of hospitals in Liaoning province，involving a total of 35 hospitals. Ten early pregnancy fetal ultrasound screening cases were randomly selected from each hospital. The specific evaluation content included the number of stored ultrasound images per examination，the completion rate of standard sections，image quality scores for each examination and individual images，excellence rates and pass rates. The assessment results were statistically described and analyzed.Results:The accuracy rate of PSAIS classification reached to 88.25%（706/800）；the ICC between PSAIS and expert ratings was 0.978. PSAIS analysis showed that the time for each early pregnancy fetal ultrasound examination was less than the average time required by experts（3.0 s vs. 63.5 s， P<0.05）.Quality control checks revealed statistically significant differences among hospitals at various levels in terms of the number of stored images，completeness of sections，scoring，excellence rate，and pass rate（all P<0.05）. Among them，tertiary hospitals performed better than secondary hospitals，private hospitals outperformed public ones，and specialized hospitals excelled over general hospitals. In the evaluation of section standardization，the fetal median sagittal plane had the highest completion rate［99.43%（348/350）］，while the umbilical cord abdominal wall entry transverse plane scan had the lowest completion rate［31.14%（109/350）］. Conclusions:PSAIS can reliably assess the quality of early pregnancy fetal ultrasound examinations，significantly enhancing the efficiency of quality control. There are noticeable differences among hospitals at various levels in Liaoning Province regarding the image quality of early pregnancy fetal ultrasounds，and there are also some common issues in standardizing sections. The quality of obstetric ultrasound images still needs further improvement.
