Interobserver reproducibility of intracranial anatomy assessment during second trimester sonographic scan

Abstract

To evaluate the reproducibility of mid-trimester intracranial anatomy assessment. Women undergoing mid-trimester scan (between 18th to 25th gestational weeks) for fetal anatomy assessment were included in the study. Measurements of lateral ventricle (LV), cisterna magna (CM) and transverse cerebellar diameter (TCD) were obtained for reproducibility analyses. Inter-observer reproducibility analysis was made with correlation coefficients. A total of 162 women were included in the analysis while one woman was excluded from the study due to diagnosis of vermian agenesis. Correlation coefficient (CC) of cisterna magna measurements have shown weak to moderate interclass correlation (r=0.28, P=0.001). Measurements of lateral ventricle have shown moderate to strong interclass correlation (r=0.73, P<0.0001). Transverse cerebellar diameter measurements have shown the best interclass correlation (r=0.88, respectively P<0.0001). Mean difference between different observers were-0.4 mm (95% CI:-3.0 to 2.1 mm), 0.7mm (95% CI:-1.0 to 2.5 mm) and 0.53 mm (95% CI:-2.5 to 3.6 mm) for measurements of cerebellum, lateral ventricle and cisterna magna, respectively. When grouped according to BMI (Over 30 kg/m2 and lower than 25 kg/m2) and presentation (cephalic and breech), there were no differences between groups regarding the prevalence of an absolute difference gre ater than 1 mm between paired measurements by different observers. Measurements of TCD, LV and CM during mid-trimester scan have good interobserver reliability with the exception of CM measurements. Methods used for measuring these structures have shown good consistency between different BMI categories and different fetal presentations during ultrasound scans.