Clinical value of fetal facial profile markers during the first trimester

Abstract

Objectives To study the correlations between facial profile markers and crown-lump length (CRL) in a Chinese population, and to evaluate the clinical value of these markers for abnormal fetuses during the first trimester (11 to 13+6 gestational weeks). Methods The facial profile markers were as followings: inferior facial angle (IFA), maxilla-nasion-mandible (MNM) angle, facial maxillary angle (FMA), frontal space (FS) distance and profile line (PL) distance. These markers were measured in facial mid-sagittal section through ViewPoint 6 software. The diagnostic value of these markers for abnormal fetuses was assessed by receiver operating characteristic (ROC) curves. Results According to the first-trimester scanning (FTS) and follow-up, 31 fetuses were enrolled in the abnormal group, including 14 cases of trisomy 21, 7 cases of trisomy 18, 10 cases with cleft lip and palate (CLP), and 1000 normal fetuses were selected. Among the normal fetuses, the IFA, FS distance and PL distance had negative correlations with CRL. The MNM angle and FMA had positive correlations with CRL. The mean IFA values for fetuses with trisomy 21 and trisomy 18 were 74.11° (standard deviation (SD) 7.48) and 69.88° (SD 7.08), respectively, which were significantly smaller than the normal fetuses (p = 0.013; p = 0.003). The mean MNM angle of fetuses with trisomy 18 and CLP were 6.98° (SD 2.61) and 9.41° (SD 2.57), respectively, which were significantly greater than the normal fetuses (p = 0.005; p < 0.001). The mean FMA values of trisomy 18 fetuses were 63.95° (SD 4.77), which was significantly smaller than the normal fetuses (p < 0.001). The mean FS distance of CLP fetuses was -0.22 mm (SD 1.38), which was significantly smaller than the normal fetuses (p < 0.001). The mean PL distance of trisomy 21, trisomy 18 and CLP fetuses were 2.89 mm (SD 0.41), 2.91 mm (SD 0.56) and 2.71 mm (SD 0.37), respectively. The difference with the normal fetuses had no statistical significance (p = 0.56; p = 0.607; p = 0.54). Conclusions Fetal facial profile markers had excellent correlations with CRL during the first trimester. IFA had certain clinical significance in detecting trisomy 21. FMA, IFA and MNM angle were reliable indicators for screening trisomy 18. The abnormal MNM angle and FS distance could be used as sensitive indicators for CLP. However, PL distance was not the best markers for trisomy 21, trisomy 18 and CLP.

A novel artificial intelligence model for fetal facial profile marker measurement during the first trimester

“…At present, studies on these facial markers mainly focus on the second and third trimester, we previously demonstrated the feasibility of measuring these markers during the first trimester, with excellent intra- and inter-operator consistency [ 18 ]. We further concluded that these markers have certain diagnostic value for fetal abnormalities [ 24 ]: IFA had a certain value in the diagnosis for trisomy 21 and trisomy 18; FMA had the excellent accuracy in detecting trisomy 18; MNM angle and FS distance were reliable indicators for screening CLP, which was consistent with this study.…”

“…However, manual measurement is laborious and complex, and identifying a technique that can precisely and swiftly measure fetal facial markers is of significant clinical value. In this study, the average inference speed achieved by our model was 0.76 s per image, significantly outperforming manual measurements, which required approximately 2 min per image [ 24 ]. Our AI-based approach for measuring facial markers can accelerate the measurement process by a factor of approximately 120.…”

A novel artificial intelligence model for fetal facial profile marker measurement during the first trimester

“…At present, studies these facial markers mainly focus on the second and third trimester, we previously demonstrated the feasibility of measuring these markers during the rst trimester, with excellent intraand inter-operator consistency. 18 We further concluded that these markers have certain diagnostic value for fetal abnormalities 20 : IFA had a certain value in the diagnosis for trisomy 21 and trisomy 18; FMA had the excellent accuracy in detecting trisomy 18; MNM angle and FS distance were reliable indicators for screening CLP, which was consistent with this study.…”

“…In this study, the average inference speed achieved by our model was 0.76 sec per image, signi cantly outperforming manual measurements, which required approximately 2 min per image. 20 Our AI-based approach for measuring facial markers can accelerate the measurement process by a factor of approximately 120. These ndings suggest that the AI model for measuring facial markers can facilitate manual measurements, improve work e ciency of sonographers and accelerate early clinical evaluation of fetal prognosis.…”