EVALUATION OF THERMAL DILUTION TECHNIQUE FOR MEASUREMENT OF STEADY AND PULSATILE FLOWS

The Stewart-Hamilton equation or its thermal analog, which is used to compute cardiac output, is based upon the assumptions of well mixed, steady flow and a mean flow sampling technique. When thermal dilution measurements are made in the aorta, the conditions required by these assumptions are not strictly met, since the flow is pulsatile and a local point sampling technique is employed. The purpose of the investigation reported here was to examine in a model system the accuracy of the thermal dilution technique under pulsatile flow conditions likely to exist in the human aorta, and in particular the influence of sensor positioning. The results of this model study showed the technique to be accurate to within 10 per cent for all flow conditions studied, as long as the fluid was well mixed at the sampling site. The ratio of indicated to actual mean flowrates was not affected by the radial location of the sensor, mean Reynolds number, frequency parameter or amplitude ratio. However, a buoyancy effect was observed which gave rise to significant errors in flowrate determination under certain conditions which may be representative of some applications of the local thermal dilution technique in short segments of small diameter vessels.