Nicholas Hilliard MBBChir, Presenter: Nothing to Disclose

Rebecca Baker, Abstract Co-Author: Nothing to Disclose

Andrew Patterson PhD, Abstract Co-Author: Nothing to Disclose

Martin John Graves BA, Abstract Co-Author: Nothing to Disclose

Christoph Lees, Abstract Co-Author: Nothing to Disclose

Patricia Ai Khoon Set MBBS, Abstract Co-Author: Nothing to Disclose

David John Lomas MD, Abstract Co-Author: Nothing to Disclose

Hypothesis: Hydrographic MR imaging can provide a rapid non-invasive in-vivo estimate of amniotic fluid volume (AFV). Background: Current ultrasound based methods, such as amniotic fluid index and single deepest vertical pocket, are indirect measures of AFV and are known to have limitations, making their use in routine management and research controversial.

23 women with healthy singleton preganancies between 28 and 32 weeks gestation were consented for MR examinations of the gravid uterus using a 1.5T MRI system. Two breath-hold techniques were used: (1) 2D 5mm thick section FIESTA, surface array coil (2) 2D 200mm thick section FSE TE 800ms, integrated volume body coil A reference fluid volume of 50mls normal saline was positioned anterior to the abdominal wall and within the field of view of (2). Manual planimetry was used to outline all of the amniotic fluid demonstrated on each 5 mm section of (1), which were summed to provide the reference standard for AFV. Manual regions of interest were used to outline the reference volume and amniotic fluid sac on (2). Using the signal area product, the volume of the amniotic fluid was estimated. The maximum values from the 5 acquisitions were compared with the reference planimetry results using a non-parametric Spearman's rank correlation.

Fluid volumes between 146 and 884 mls were found on planimetry. High inter-rater agreement was noted for both the methods (ICC=0.961 and 0.997). The rank order correlation between the planimetry and the hydrographic method was highly significant (r=0.864, p<0.001). A linear fit equation of y=0.6083 + 163.05mls was obtained, with planimetry defined as the independent variable. This relationship suggests that the inclusion of fetal fluid structures is likely to bias the results positively at lower AFV, and the inhomogeneity of B1 excitation is likely to bias the results negatively at larger AFV.

This initial study indicates that it is possible to estimate AFV with MRI using a rapid hydrographic technique, based on single thick slab acquisitions.  Further optimisation for fetal fluid structures, RF inhomogeneity, as well as data at different gestational ages will be required.

A rapid MR hydrographic based estimate of amniotic fluid volume may allow for improved pregnancy management, and new research into fetal outcomes.

Hilliard, N, Baker, R, Patterson, A, Graves, M, Lees, C, Set, P, Lomas, D, Amniotic Fluid Volume Estimation by MR Hydrography.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14009446.html