The aim of the present study was to design, develop and examine a portable
and low cost magnetometer for cardiac imaging. Magnetocardiography (MCG)
involves capturing magnetic field maps (MFM's) of current distributions resulting
from cardiac action potentials. Particular emphasis was given to the diagnostic
performance of distinguishing cardiac patients, mainly with ischemic heart disease,
from healthy individuals with the aim of potentially expanding the clinical utility.
The current magnetocardiography is known as a superconducting quantum interference
device (SQUID). It was first used in 1970 in a magnetically shielded room
and the device produced an MCG waveform that could be visualised just as clearly
as the conventional electrocardiogram. This popularity is largely attributable to
the appearance of unshielded SQUID devices in the late 1990s. However, continued
reliance on large immovable equipment that is expensive to maintain, has
limited the clinical application of MCG to a small but expanding group of research
clinicians worldwide. Despite this, a large body of evidence suggests that an unshielded
MCG device is a potentially powerful tool for detecting a multitude of
cardiac abnormalities in the human heart. Over the course of this research, I have
constructed an induction coil magnetometer that is not only portable, low cost
and non-invasive, but also reaches into the sensitivity range needed for magnetocardiography
and produces magnetic field maps that are medically diagnosable.
Therefore, this technology provides a real alternative to other costly diagnostic
tools, such as Superconducting Quantum Interference device-SQUIDs.
While MCG sensors are an established diagnostic tool to diagnose cardiac conditions,
for other groups around the world [1], this device is a newly developed
apparatus using a less complicated design. This study determines the extent to
which the device can provide a clear image across test subjects with a range of age
and genders and to identify which parameters from the MCG data are markers for
ischemic heart disease when compared with healthy controls.
A pilot study was conducted using 40 patients with ischemia and 20 with post
myocardial infarction (MI) as well as 60 healthy age matched controls to evaluate
the capability of MCG to detect myocardial ischemia. To identify the optimal
recording of MCG, we looked at 55 patients with ischaemic heart disease and
10 post MI patients. In addition, 49 healthy age matched controls underwent
examinations. In the MCG pilot study, we defined standard features or \lobes" of
MCG signals, which are common to all healthy normal subjects. For the purpose of
evaluating the diagnostic performance of the MCG these are referred to as cursors,
and consist of 6 parameters that can be estimated, QR and RS, T-wave (T1-T4), as
well as the RT interval and zero crossings (ZCs). We also consider QR length and
RS length, T length, and angles and genders. They were examined in the detection
of myocardial ischemia in both the patient group and the healthy controls. The
results showed that MCG is able to distinguish patients from healthy regardless
of the presence or absence of a history of remote MI.
Stepwise regression was performed using the Forward Logistic Regression-LR method.
Using this approach, any variable with p > 0:20 was removed at each step. Thereafter,
the two-way interactions were included in a stepwise analysis with other
significant predictors. Candidate predictors entered at step 1 include the following;
T2 corr by T2 length, QR cursor, QR angle, T1 angle, T zcMDIS, QR length
by RS length, and GENDER(1) by T stdcorrvar. The aforementioned parameters
are spatial and temporal diagnostic variables of the MCG that we used in o
ur study to distinguish healthy from patient. T1 is the magnetic Field Map (MFM)
at the beginning of depolarization and T2 is the MFM at the peak of the depolarsation.
Length, angle and correlation of each of these were measured. We also
defined some other parameters such QR that is repolarisation at the peak of Rwave.
More definitions are detailed in chapter 6. As a result of the final model, we
find the following interactions are significant predictors: QR length by RS length,
GENDER(1) by T stdcorrvar and T2 corr by T2 length.
In conclusion, 15 channel- MCG can detect myocardial ischemia at rest and supine
position in an unshielded environment. In MCG several factors can be used to
distinguish unhealthy from healthy: QR, RS length and T-wave correlations. The
interactions between some of the parameters can also produce significant predictions.
We also found that T-wave changes during the period of time corresponding
to T1 to T4 are very useful in detecting abnormalities. Furthermore, the T-wave
parameters such as angles and lengths may contain information on separate physiological
stages of the ischemia development. This will be subject to future research.
