Thermal dilution: instrumentation with thermistors

¹ Department of Electrical Engineering, Northwestern University, Evanston, Northwestern University Medical School, and Division of Cardiology, Children's Memorial Hospital, Chicago, Illinois

Based on the steady-state small-signal equivalent circuit parameters for a self-heated thermistor, the relationship of flow versus temperature sensitivity is found to be linear with the power dissipated in the thermistor. For the thermistors considered (YSI 520 and 512 series) flow sensitivity is negligible at power inputs below .01 mw. This power applied to a thermistor incorporated into a balanced Wheatstone bridge provides an output signal of approximately 150 µv for a .1 C temperature change. Amplifier requirements for thermal dilution are: gain of 10⁴, a frequency response of 25 cycle/sec or greater, and a peak noise referred to input of less than 2 µv. A calibration procedure using a fixed resistance bridge unbalance and calculations from measured thermistor parameters is discussed. The procedure is independent of changes in amplifier and recorder gain, or changes in bridge voltage and ambient temperature. The effect of thermistor aging on the procedure is considered and found to be quite small. A bridge circuit incorporating variable bridge voltage is described and the application of the variable bridge voltage to the in vivo estimation of the thermistor dissipation constant and flow effects are discussed. Intravascular temperature measurements in man and in the anesthetized dog indicate that changes in thermistor resistance periodic with pulse are primarily the result of intermittent contact of the thermistor with the vessel wall, and due to changes in both temperature and dissipation constant.

Submitted on June 12, 1964