METHODS OF MEASURING BLOOD FLOW 



'3'5 



tion, lower flow signals are permissible. The magnetic 

 field strength and the magnet size may be greatly- 

 reduced, and the method can therefore be applied to 

 very small vessels. Since miniaturization of the 

 magnet-sleeve-electrode assembly is possible, the de- 

 vices may even be adapted for chronic implantation. 

 Finally, any spurious potentials which change in 

 time at a much lower frequency than that of the 

 carrier used can be excluded from registration by 

 appropriate circuitry of the amplifier and demodula- 

 tor (123). Usually a carrier frequency of a few hundred 

 cycles per second will be high enough to cancel cardiac 

 action potentials. 



A sufficiently high carrier frequency should be 

 chosen for a more important reason (123). If A/ is the 

 highest frequency of the flow oscillations to be re- 

 corded and / is the carrier frequency, a pass band 

 reaching from (/ — Af) to (/ + Af) has to be amplified 

 while by the demodulation and output filtering the 

 carrier is suppressed and a filtered output signal of a 

 frequency range from o to Af is obtained for registra- 

 tion. According to generally known principles,/ must 

 be higher than lAf. In case of flow recording, Af is 

 usually about 50 cps (35); in special cases, Af may be 

 higher, say up to about 100 cps. Therefore, / should be 

 above 200 cps, i.e., 300 to 500 cps. On the other hand, 

 / should not be higher than necessary because par- 

 ticular difficulties (see below) will increase with rising 

 frequency. A carrier frequency of about 400 cps will 

 permit reaching an adequate frequency response of 

 the flow signal without undesired nonmodulated sig- 

 nals such as those of the ECG. Standard line current 

 (J = 60 or 50 cps), which was used in the earlier de- 

 velopmental stages of the a-c method, may be suc- 

 cessfully used to energize the magnet when a smaller 

 frequency range is sufficient (89, 1 1 1, 132). 



A particular difficulty hampering the a-c sine-wave 

 procedure has been hardest of all to overcome. The 

 electrode leads, the electrodes, and the vessel segment 

 lying between them form a transformer loop in which 

 an a-c voltage is induced by the alternating magnetic 

 field. This spurious voltage, which is commonly called 

 "transformer component" or "transformer emf," is 

 proportional in strength to the rate of change of the 

 magnetic field strength (dB/dt) and also depends on 

 the configuration of the effective transformer loop. It 

 follows from equation 15 that dB/dt = .Boar cos cat, 

 and the transformer component will be E t = 

 A'Sow cos o)<, where A" = coefficient related to the loop 

 configuration. Thus we get the actual a-c voltage £,■ 



fed into the amplifier input (66, 80) as the sum: 



E r E f +E t' 



B-IO' B (DV smut + Kwcosut) volts. 



O A 



(17) 



Obviously, E, leads in phase by 90 and will rise in 

 amplitude with increasing frequency of the magnet 

 current. Additional spurious a-c voltage may also be 

 created by the stray capacitance of the magnet coils 

 on the one hand and the vessel, electrodes, and leads 

 on the other; it can be minimized by electrostatic 

 shielding of sleeve and leads and by using a balanced 

 push-pull or differential amplifier with high common- 

 mode rejection. 



Various designs have been worked out [for a survey 

 see (123)] for eliminating or at least minimizing the 

 transformer emf. The most important of these are as 

 follows: a) A special variable-phase transformer is fed 

 by the magnet current; the output of this transformer 

 is arranged in series with one of the electrode leads so 

 that cancellation of the "transformer component" is 

 wrought by an opposite-phase emf delivered by the 

 phase transformer (Kolin). Complete cancellation is 

 impossible because of wave form distortion due to non- 

 linearity of magnetic-core material, b) Cancellation 

 voltage is derived from a special pickup coil wound 

 around the magnet core (18, 112, 113). Complete 

 cancellation is possible with this refinement by careful 

 adjustment of a potentiometer, c) A split-lead method 

 (28) can be used. It is a modification of b in which one 

 electrode lead is split; the halves are placed on either 

 side of the magnet core and are joined by a potenti- 

 ometer. This arrangement also makes cancellation 

 possible, d) An auxiliary pickup coil can be arranged 

 at the sleeve (80) in external-magnet devices, e) Care- 

 ful orientation of both electrode leads can form a non- 

 inductive loop (66, 80, 82). This technique is mainly- 

 applied in magnet-sleeve units. It permits approxi- 

 mate cancellation. Complete cancellation can be 

 achieved by additional phase-sensitive demodulation 

 (82). /) Demodulation with phase discrimination 

 should, theoretically, permit separating E t from E/ to 

 obtain the latter only, since they differ in phase by 90°. 

 Simple demodulators consisting of half-wave or full- 

 wave rectifiers cannot provide any- separation of sig- 

 nals by their phase. Kolin (77) first applied optical 

 phase discrimination on the screen of an oscilloscope. 

 James (66) suggested electronic phase discrimination 

 in a-c flowmeters. Continuous phase-sensitive de- 

 modulation or discontinuous discrimination by gating 



