Multilevel modeling of fetal and placental growth using echo-planar magnetic resonance imaging

J Soc Gynecol Investig. 2001 Sep-Oct;8(5):285-90. doi: 10.1016/s1071-5576(01)00126-5.

Abstract

Objective: To quantify longitudinal increases in fetal, fetal liver, and fetal brain volume using echo-planar magnetic resonance imaging and to quantify the results using appropriate statistical modeling.

Methods: Fifty-six singleton fetuses were studied using echo-planar (snap-shot) magnetic resonance imaging, between 19 weeks and term. They were assessed at a variety of different gestations and on a different number of occasions, thereby requiring multilevel statistical modeling to analyze the pattern of fetal growth.

Results: Fetal volume varied according to the following equation: square root (radical) [fetal volume]=-37.71+2.17 x gestational age (GA)-0.004 x GA(2). The equation for fetal liver volume was radical[fetal liver volume]=9.47+0.56 x GA-0.02 x GA(2), for fetal brain volume was radical[fetal brain volume]=15.50+0.69 x GA-0.014 x GA(2), and for placental volume radical[placental volume]=28.54+0.95 x GA-0.039 x GA(2), where GA is the gestational age in weeks -30.

Conclusion: The assessment of fetal, fetal organ, and placental volume was feasible using echo-planar magnetic resonance imaging from 20 weeks to term. Multilevel statistical modeling can be applied to analyze sets of data with different measurements on different occasions. This information is useful clinically to assess abnormal fetal growth.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Brain / embryology
  • Echo-Planar Imaging
  • Embryonic and Fetal Development / physiology*
  • Female
  • Humans
  • Image Processing, Computer-Assisted
  • Infant, Newborn
  • Liver / embryology
  • Longitudinal Studies
  • Male
  • Models, Statistical
  • Placenta / physiology
  • Placentation*
  • Pregnancy
  • Prospective Studies