To investigate the current quality status of prenatal ultrasound screening for fetuses at 11-13⁺⁶ weeks of gestation in Jiangxi Province, analyze the technical issues and shortcomings in early pregnancy ultrasound screening images, propose targeted capacity improvement strategies, and provide a scientific basis for optimizing early pregnancy ultrasound screening services in Jiangxi Province.

A total of 49 659 ultrasound screening images of fetuses at 11-13⁺⁶ weeks of gestation, obtained from 558 ultrasound physicians across 106 medical institutions in Jiangxi Province between January and December 2024, were collected via the "intelligent ultrasonic quality control system". AI-generated quality scores for these images were analyzed. The ultrasound screening views included two quality-controlled fetal views (crown-rump length [CRL] and nuchal translucency [NT] views) as well as multiple non-quality-controlled views. Independent samples t-test and one-way analysis of variance (ANOVA, F-test) were used to compare quality scores of early pregnancy ultrasound screening images across different types of medical institutions, categories of ultrasound physicians, and prefecture-level cities. Additionally, differences in quality confidence coefficients for structural markers in different views were analyzed.

The quality scores of early pregnancy ultrasound screening images in maternal and children hospitals were significantly higher than that in general hospitals [(78.8±18.0) vs (77.5±18.0), t=6.38, P＜0.001]. Tertiary hospitals had significantly higher image quality scores than secondary hospitals [(81.0±14.7) vs (76.2±20.5), t=29.38, P＜0.001]. Image quality scores decreased progressively from provincial-level hospitals to municipal-level hospitals and then to county-level hospitals [(83.2±13.8) vs (80.4±14.9) vs (76.7±20.1), F=370.70, P＜0.001]. Higher physician titles and longer ultrasound practice experience were associated with higher quality scores for all views. Physicians with a master's degree or above achieved significantly higher quality scores for all views compared to those with a bachelor's degree or junior college education (all P＜0.001). Significant differences in image quality scores were observed among 11 prefecture-level cities in Jiangxi Province (all P＜0.001). Both CRL and NT screening views contained 12 structural markers each, and there were statistically significant differences in quality confidence coefficients for structural markers across all views (all P＜0.001). The main quality issues of the images included: unclear display of important structural markers, insufficient magnification, and failure to avoid unwanted structures.

Significant differences exist in the quality of prenatal ultrasound images for fetuses at 11-13⁺⁶ weeks of gestation across different types of medical institutions, ultrasound physicians, and prefecture-level cities in Jiangxi Province. To improve the quality of early pregnancy ultrasound screening in Jiangxi Province, priority should be given to general hospitals, municipal/county-level hospitals, and ultrasound physicians with lower titles, lower academic qualifications, and less clinical experience. Training and education should focus on quality control standards, such as the clarity of structural markers, appropriate magnification, and avoidance of unwanted structures. This will enhance the examination proficiency of ultrasound physicians at all levels and types, thereby improving the overall capacity of early pregnancy ultrasound screening in Jiangxi Province.