
Page
20
of
23
Document number 291945 V5
Basic configuration for non-contact ECG measurement including capacitively-
coupled DRL”
circuit.
In non-contact ECG measurement there is
– by definition - no skin contact, and thus no direct
connection can be made between the subject’s body and the system ground. Some other
method of reducing the power line noise is therefore required to enable the ECG signal to be
extracted reliably and accurately. One such method utilises an approach very similar to the
Driven Right Leg (DRL) system that is used for the same purpose in conventional ECG. In
conventional ECG the DRL signal is coupled directly to th
e patient’s skin; in non-contact
ECG it is coupled capacitively to the body, through clothing, via a piece of conductive material
placed
– for instance – on the seat or back of a chair. Capacitive coupling of DRL signals is
described by Lim et al 1 and Lee et al2.
1.
Settling time
– When a subject first sits in the chair and leans against the EPIC sensors,
the changes in electric potential will normally send both the sensors and the DRL circuit into
saturation. Because the system contains some large impedances, and hence has some very
long RC time constants, settling times of tens of seconds can be needed before a clean ECG
signal is seen. During this period the signal can either appear very noisy, or be virtually flat,
depending on whether one or both se
nsors, or the DRL circuit, are “railing”. The subject
should sit still during this time and wait for the circuit to settle, since continually adjusting
position will only make matters worse. Settling times can sometimes be reduced by turning
off the power to the demo box for a few seconds.
2.
Clothing
– Best results have been obtained when the material between the sensors
and the skin is one or two layers of cotton material. This is therefore recommended as a
starting point for system evaluation. Signals have also been measured successfully through
other materials, including a wool-mix sweater and a polyester fleece in addition to two layers
of cotton material. Examples are shown in figures 6 and 7. If the key parameter of interest
is the “R-R” interval, adjustment of filter settings to reduce or re-centre the signal bandwidth
can give significant improvement in signal quality.
EPIC Sensors in
contact with clothing
Conductive fabric in contact with
clothing, e.g. on chair seat
Output
–
EPIC demo
box