Accuracy of volume measurement by conductance catheter in isolated, ejecting canine hearts

We evaluated the accuracy of the recently reported technique of estimating intraventricular volume by measurement of intracavitary electrical conductance in six isolated, ejecting, canine left ventricles. Left ventricular volumes were measured directly by a previously validated servosystem that employed an electroconductive balloon placed in the left ventricular cavity. The volume measured continuously by the balloon method (V(bal)) was compared with that estimated by the conductance method (V(cath)). For this test, the hearts were made to eject and fill physiologically by the use of a previously described computer-simulated arterial loading system. Complex ejection and filling patterns were created by stimulating the atrium mechanically, which resulted in irregular arrhythmatic contractions spanning a wide range of volumes. We found that there was a highly linear relationship (r² = .982 ± .014) between V(bal) and V(cath): V(cath) = 0.82 (± .05) V(bal) + 26.7 (± 11.8) ml. Despite the wide variation in the offset term of this relationship among the different hearts, the offset within a given heart was predicted within 3.5 ml by a previously detailed 'dilution' method that is applicable to the heart in situ within a closed thorax. Thus, since the offset term is obtainable in situ, the conductance method provides a signal that is proportional to the actual volume. To determine whether right ventricular volume influenced the accuracy of left ventricular measurement, we compared the relationship between V(cath) and V(bal) obtained with right ventricular volumes of 0 and 30 ml. Increasing the right ventricular volume shifted the relationship upward by less than 3 ml in the working range. Finally, with the left ventricle constrained to contract isovolumetrically, there was less than ± 2 ml variation about a mean value in the conductance signal, indicating that the geometric rearrangements occurring during contraction do not significantly influence the accuracy of this method. We conclude that the conductance method of left ventricular volume measurement provides a continuous signal that is proportional to the actual chamber volume and may therefore prove to be useful in the assessment of ventricular performance in patients and in laboratory animals.