Automatic capture verification can prolong pulse generator longevity and in
crease patient safety. However, the detection of evoked response following
pacing is complicated due to afterpotentials caused by polarization of elec
trodes. This study describes a new capture verification scheme, which neutr
alizes the charges between the pacing electrodes. The hypothesis of the cha
rge-neutral sensing is that the afterpotentials in the ring and the tip are
opposite in polarity when pacing in a bipolar mode between ring and tip. S
umming the unipolar signals sensed at the tip and the ring should effective
ly cancel the afterpotentials. This scheme was implemented in an external c
omputer based system and tested during pacemaker implant/replacement on 23
patients during VVI pacing (17 acutely implanted leads and 6 chronic leads)
. Surface ECG was recorded to provide a marker for capture and noncapture.
The pacing voltage was gradually decreased until a noncapture beat was note
d. To avoid fusion beats, the pacing rate was programmed similar to 50% hig
her than the intrinsic rate. The evoked response was high pass filtered and
the integral average was calculated for both capture and noncapture beats.
The system signal to noise ratio (SNR) was expressed as ratio of the minim
um integral average of all capture beats to the maximum integral average of
all noncapture beats. The system SNR was 8.6 +/- 1.3 (mean +/- S.E.M; rang
e 1.5-22.8), indicating that the charge-neutral sensing method has, on aver
age, a ninefold safety margin in providing capture verification. Further, e
valuation is needed to fully assess this feature in patients with chronic l
eads.