Reference ranges of fetal cisterna magna length, area and volume measurements by three-dimensional ultrasound

Objetivo: Determinar valores de referencia para o comprimento, area, volume bidimensional (2D) e tridimensional (3D) da cisterna magna fetal por meio da ultrassonografia tridimensional (US3D). Metodos: Realizou-se um estudo de corte transversal com 224 gestantes normais entre 17 a 29 semanas completas. A mensuracao da area e comprimento da cisterna magna fetal foi realizada no plano axial (A) do modo multiplanar, ao nivel do plano de medida do diametro transverso do cerebelo. A mensuracao do volume da cisterna magna fetal 2D foi realizada no modo multiplanar por meio do produto de tres medidas, sendo duas no plano axial e uma no plano sagital, pela constante 0,52. A medida do volume 3D foi realizada no plano axial padrao de medida do diametro transverso do cerebelo, por meio do metodo Virtual Organ Computer-aided AnaLysis (VOCAL) com angulo de 30º. Para a determinacao de valores de referencia, realizaram-se regressoes polinomiais com ajustes pelo coeficiente de determinacao (R2). A confiabilidade e concordancia foram realizadas pelo coeficiente de correlacao intraclasse (CCI) e limites de concordancia dos graficos de Bland-Altman. Para se avaliar a correlacao entre os volumes 2D e 3D, utilizou-se o coeficiente de correlacao de Pearson (r). Resultados: A media do comprimento, area, volume 2D e volume 3D da cisterna magna fetal variou de 0,50±0,10 a 0,79±0,18 cm; 0,95±0,18 a 3,09±0,62 cm2, 0,71±0,19 cm3 a 4,18±0,75 cm3 e 0,73±0,25 cm3 a 3,79±1,10 cm3 entre 17 a 29 semanas; respectivamente. Observou-se uma a baixa correlacao do comprimento e uma alta correlacao da area, volume 2D e volume 3D com a idade gestacional (IG): -1,426+0,1551xIG-0,0027xIG2 (R2=0,36); -3,055+0,2157xIG (R2=0,70); -1,918+0,0284xIG+0,0065xIG2 (R2=0,67) e-1,918+0,0284xIG+0,0065xIG2 (R2=0,67); respectivamente. Observou-se boa confiabilidade e concordancia intraobservador para o comprimento e area com CCI=0,86 e limites de concordancia 95% (-24,3; 24,4) e CCI=0,91 e limites de concordancia 95% (-26,6; 35,6); respectivamente. Observou-se boa confiabilidade e concordancia interobservador para o comprimento e area com CCI=0,64 e limites de concordancia 95% (-43,1; 40,3) e CCI=0,82 e limites de concordancia 95% (-51,0; 49,2); respectivamente. Observou-se boa confiabilidade e concordancia intraobservador para o volume 2D com CCI=0,89 e limites de concordancia 95% (-52,0; 51,8). Observou-se baixa confiabilidade e concordancia interobservador para o volume 2D com CCI=0,64 e limites de concordancia 95% (-110,1; 84,6). Observou-se boa confiabilidade e concordancia intraobservador para o volume 3D com CCI=0,92 e limites de concordancia 95% (-49,7; 48,4). Observou-se uma baixa confiabilidade e concordancia interobservador para o volume 3D com CCI=0,58 e limites de concordancia 95% (-114,0; 80,2). Observou-se uma boa correlacao entre os volumes 2D e 3D (r=0,879). Conclusoes: Determinou-se valores de referencias para o comprimento, area, volume 2D e volume 3D da cisterna magna fetal, sendo que todos os metodos apresentaram boa confiabilidade e concordancia intraobservadorObjective To determine reference values for the length, area, two-dimensional volume (2D) and three-dimensional (3D) volume of the fetal cisterna magna through threedimensional ultrasonography (3DUS). Methods: A cross-sectional study with 224 healthy pregnant women between 17 to 29 completed weeks was conducted. The area and length measurements of the fetal cisterna magna were performed in the axial plane (A) of the multiplanar mode, on the level of cerebellar transverse diameter measurement plan. The 2D volume measurement were performed through the multiplanar mode by the multiplication of three diameters by the 0.52 constant; two in the axial plan and one in the sagittal plan. Measurement of the volume of the fetal cisterna magna was performed in the standard axial plane of the transverse diameter of the cerebellum through the VOCAL method with a 30 ° angle. For the determination of reference values, polynomial regressions with adjustments by the coefficient of determination (R2) were held. The reliability and agreement were made by the intraclass correlation coefficient (ICC) and limits of agreement of the Bland-Altman. For the evaluation of a correlation between 2D and 3D volumes, the Pearson correlation coefficient was used (r). Results: The mean length, area, 2D and 3D volume of the fetal cisterna magna ranged from 0.50±0.10 to 0.79±0.18 cm; 0.95±0.18 to 3.09±0.62 cm2, 0.71±0.19 cm3 to 4.18±0.75 cm3 and 0.73±0.25 cm3 to 3.79±1.10 cm3 between 17 and 29 weeks, respectively. It was observed a low correlation of the lenght and a high correlation of the area, 2D and 3D volume with the gestational age (GA): -1.426+0.1551XGA-0.0027xGA2 (R2=0.36); -3.055+0.2157xGA (R2=0.70); - 1.918+0.0284xGA+0.0065xGA2 (R2=0.67) and -1.918+0.0284xGA+0.0065xGA2 (R2=0.67); respectively. It was observed good reliability and intraobserver agreement for length and area with ICC = 0.86 and limits of agreement 95% (-24.3, 24.4) and ICC = 0.91 and limits of agreement 95% (-26.6; 35.6), respectively. It was also observed good reliability and interobserver agreement for length and area with ICC = 0.64 and 95% limits of agreement (-43.1, 40.3) and ICC = 0.82 and 95% limits of agreement (-51.0, 49.2), respectively.). It was observed good reliability and intraobserver agreement for the 2D volume, with ICC = 0.89 and 95% limits of agreement (-52.0, 51.8). There was low interobserver reliability and agreement for the 2D volume, with ICC = 0.64 and 95% limits of agreement (-110.1, 84.6). It was observed good reliability and intraobserver agreement, with ICC = 0.92 and 95% limits of agreement (-49.7, 48.4). There was low interobserver reliability and agreement, with ICC = 0.58 and 95% limits of agreement (-114.0, 80.2). It was observed a good correlation between 2D and 3D volumes (r=0.879). Conclusions: Reference values for the length, area, 2D and 3D volumes of fetal cisterna magna were determined and showed good agreement and intraobserver reliability in all methods.BV UNIFESP: Teses e dissertaçõe