Electrocardiography (ECG) is the gold standard for rhythm assessment; however, fetal ECG (fECG) is not used clinically for assessment of fetal rhythm due to its limited signal quality and success rate. The subject of this chapter is a relatively new technique, fetal magnetocardiography (fMCG), FMCG is the magnetic analogue of fECG; that is, the same cardiac currents that generate the fECG also give rise to magnetic signals that comprise the fMCG. The fMCG and fECG are both directly proportional to the net heart current; thus, they are similar in appearance and information content. The fECG, however, suffers from a low and/or variable signal amplitude, especially during the critical period of pregnancy from about 26 to 35 weeks when the fetal skin is covered by the vernix caseosa. The high electrical resistance of the fetal skin and vernix impedes the transmission of currents to the maternal surface. Magnetic signals, in contrast, do not rely on volume conduction and thus are less affected.

The last decade has seen considerable progress in demonstrating the utility of fMCG for evaluation of fetal arrhythmia. Although fMCG is not widely used, its efficacy has been acknowledged in leading clinical journals and in the first-ever American Heart Association Scientific Statement on Fetal Diagnosis and Treatment.¹ This chapter introduces readers to the basic principles, technical aspects, and clinical application of fMCG. The main objective is to describe the advantages of fMCG and the rationale for its use as an adjunct to fetal echocardiography.

Currently, echocardiography is the only method used clinically to diagnose fetal arrhythmia; however, it is important to bear in mind that the mechanical rhythm assessed by echocardiography is a surrogate for electrical rhythm and has significant limitations.

The main advantage of fMCG is its ability to provide waveform information. This is critical for accurate diagnosis of ventricular rhythms, such as premature ventricular contractions (PVCs) and ventricular tachycardia (VT). Perhaps the most compelling advantage of fMCG is its ability to assess repolarization, which is mechanically silent and therefore cannot be assessed by echocardiography. fMCG can detect QTc prolongation, T-wave alternans, ST depression, and other important phenomena associated with abnormal repolarization.

fMCG is also more precise than echocardiography. It has been shown that in some cases the mechanical rhythm does not always accurately reflect the underlying electrical rhythm,²,³ which can result in an ambiguous or inaccurate diagnosis. In addition, fMCG offers some practical advantages. It is well suited for extended, Holter-like monitoring. This makes it highly useful for assessment of complex rhythms that change over time. It is also useful for detection of brief, transient rhythms, such as brief runs of VT, which are critical to catch at an early stage.

The fECG and fMCG show similar temporal behavior; however, their signal topographies differ substantially. Electric signals are due to potential differences along the direction of the current, as described by Ohm’s law. Magnetic signals are composed of flux lines that encircle the net current, as described by Ampere’s (FIG. 2.3.1). The polarity of the signal depends on whether the flux lines are entering or exiting the body surface; therefore, the polarity of the detected signal is reversed on opposite sides of the current.

Although fMCG is an effective technique, it has significant limitations relative to postnatal ECG. The main limitation is that the signal-to-noise ratio (SNR) is exceedingly low at early gestational ages due to the small size of the developing heart. Typically, studies are performed after about 20 weeks when the success rate is high. In addition to the gestational age dependence, the rapid falloff of the signal with distance is an important consideration. Thus, the lie of the fetus, the placement of the sensors, and the body mass ratio of the mother are factors that can influence the signal quality or the timing of a study.

A further limitation is that the configuration of fMCG waveforms is ambiguous because the position and orientation of the sensors with respect to the fetus depends on fetal lie and, therefore, may differ from subject to subject. In practice, recordings are made at several positions to capture a range of waveform configurations.

The main drawback of fMCG is the high cost and complexity of the instrumentation. Until recently, the only detectors with sufficient sensitivity to record the fMCG were SQUID (Superconducting QUantum Interference Device) magnetometers (FIG. 2.3.2A). The principles of operation of SQUIDs are beyond the scope of this work and can be found elsewhere.⁴ They are superconductor devices and require liquid helium, a vanishing resource. In 2016, a SQUID fMCG system (Tristan 621/624 Biomagnetometer, Tristan Technologies, San Diego) received FDA 510(k) clearance. The combined cost of this system and a magnetically shielded room is approximately $1M.

Recently, however, there has been a breakthrough in magnetic sensor technology. The new sensors, known as an optically pumped magnetometers (OPMs; FIG. 2.3.2B), are compact and perform similarly to SQUIDs without the need for liquid helium.⁵ Their small size enables the use of person-sized shields, which allows the entire system to be portable and much less expensive. Currently, there are no commercial vendors of OPM fMCG systems, but it is likely that OPMs will replace SQUIDs in the foreseeable future and that the eventual cost of an OPM fMCG system will be several times less than that of current SQUID fMCG systems.

Recording the fMCG is relatively straightforward. Typically, a brief ultrasound examination is performed to locate the fetal heart, and the detector is placed on the maternal surface in close proximity to the fetal heart. The operator sits at a computer console outside the shielded room, and the mother is instructed to lie quietly on the patient table for the duration of the recording.

The recordings typically last 5 to 10 minutes; however, typically several recordings are taken, moving the sensor position at least once in between. For cases involving complex or intermittent arrhythmias, the recording time can exceed 1 hour. For cases involving obstetrical high-risk conditions, the recording time should be at least 20 minutes so that fetal heart rate reactivity can be assessed.

In principle, only a few channels are required to capture the information contained in the signal. In practice, however, fMCG sensor arrays may comprise a dozen or more channels to facilitate the use of signal processing, which is essential for removing interference.

The raw fMCG signals are contaminated with large interference. The main interference is the maternal MCG, which is typically larger than the fMCG unless the fetus is near term. Multichannel sensors enable the use of a signal processing technique known as spatial filtering,⁶,⁷ which is the spatial analogue of temporal filtering; that is, the filter output is a linear combination of the spatial samples. The method is able to separate the fetal signal from the maternal MCG and other interferences based on differences in their spatial characteristics. These differences result largely because the fetal heart is much closer to the sensors than the maternal heart and other sources of interference.

The signal-processed recordings are used to generate fetal heart rate and fetal activity tracings, averaged fMCG waveforms, and rhythm strips. These data are used for clinical interpretation.