The influence of fetoabdominal volume conduction on the fetal MCG

Fetal magnetocardiograms (MCG) can be measured non-invasively. From these measurements the duration of the various waves in the fetal cardiogram can be obtained as well as the fetal heart rate. Hence, the fetal MCG can, for instance, be used to classify arrhythmias. However, the amplitudes within the measured cardiogram depend on the fetoabdominal anatomy. For instance, the position of the measurement position relative to the fetal heart as well as the current distribution in the tissues surrounding the fetal heart are of influence. As this influence is not well known, the amplitudes cannot be used for diagnostic purposes. In order to use amplitudes for diagnostic purposes, models of the fetoabdominal volume conductor have to be created and tested that can be used to reconstruct the strength of the current source (i.e. the heart vector). In this paper five volume-conductor models are compared to estimate its influence on the fetal MCG and to determine whether a standardized model can be used to reconstruct the heart vector. For example, near term about two thirds of the fetuses is lying in a left occipital lateral position and one third in the right. Hence, two models, one describing the fetus in a left occipital lateral position and one mirrored in the sagittal plane, may be sufficient to reconstruct the heart vector. Simulations can provide a means to calculate the influence of the volume conductor [1]. To construct a realistically shaped model of the volume conductor, information on the fetoabdominal anatomy is needed. This information can, for instance, be obtained from ultrasound images or MR-images. In this paper the fetoabdominal anatomy is subdivided into compartments with a homogeneous conductivity. The volume conductor changes during gestation. When the fetus is covered by a layer of vernix caseosa, an electrically insulating layer, the appropriate model seems a one-compartment model with the shape of the fetus. A three-compartment model describing the fetus, the amniotic fluid, and the maternal abdomen may be appropriate, for instance, before the 28 week of gestation when the layer of vernix is not present [2], or whenever the vernix is not present. Another approach may be the half-space model, as it does not require extensive computations. Simulations can also be of help to optimize the fluxtransformer used to measure the fetal (vector) MCG.