Determination of Arterial Blood Flow by Percutaneously Introduced Flow Sensors in an External Magnetic Field, I. The Method

Transcription

1 Proceedings of the National Academy of Sciences Vol. 67, No. 4, pp , December 1970 Determination of Arterial Blood Flow by Percutaneously Introduced Flow Sensors in an External Magnetic Field, I. The Method Alexander Kolin, Julius H. Grollman, Jr., Richard J. Steckel, and Harold D. Snow UNIVERSITY OF CALIFORNIA SCHOOL OF MEDICINE, LOS ANGELES, CALIF Communicated by Georg von Bekesy, September 17, 1970 A method for determining the rate of blood flow in arteries with- Abstract. out surgical intervention is described. It is shown that a probe capable of securing such information about flow in a blood vessel can be introduced into the vascular tree by means of a needle puncture in the skin made by a no. 18-gauge hypodermic needle. A snugly-fitting guide wire is passed through the needle, which is then withdrawn. A Teflon catheter is then passed over the wire and pushed through the skin and subcutaneous tissues into the superficial branch artery from which the wire guides it deep into the aorta to a desired location. The guide wire is then withdrawn and a suitable sensing device of wire-like structure can be passed through the catheter into the major arterial trunk and, if desired, into some of its branches. Two types of sensors are used, loop-shaped ones which are used as volume-flow meters, and L-shaped probes used as volumeflow meters when they fit the artery precisely and as velometers when they do not. The sensors are used in conjunction with an apparatus consisting of a flat coil mounted over the animal and a standard electromagnetic sinusoidal flow meter channel. It has been shown1'2 that blood flow may be measured electromagnetically with extremely small sensing elements if the magnetic field required to induce the flow-signal voltage in the moving blood is generated by a magnet placed outside the animal rather than made part of the intravascular flow sensor. In those studies the flow sensors were introduced into the vascular tree, via exposed femoral or carotid arteries, through silicone rubber tubes inserted into these vessels. In the present study, the flow probes are introduced percutaneously through a no. 7 French, radio-opaque, angiographic Teflon catheter, which is the size normally employed for human cardiac catheterization (2.3-mm outer diameter, 1.6-mm inner diameter). Smaller catheters (no. 5 French) can be used for this purpose, and probes can be made to pass through them and even through smaller sizes. Excessive reduction in probe diameters leads, however, to problems due to vibrations of the probe tip, small electrode size, and hazard of injuring the arterial wall. We believe that the present dimensions represent a satisfactory compromise. The Apparatus. Fig. 1 shows the setup for electromagnetic determination of blood flow by means of intravascular electromagnetic flow sensors, introduced 1769

2 1770 PHYSIOLOGY: KOLIN ET AL. PROC. N. A. S. s' s. FIG. 1. The setup for blood 2 flow determination by means of catheter sensors used in conjunction with an external magnetic field. X: x-ray tube enclosure. G I: Image intensifier. T: "Floating table-top, movable in X and Y directions in the horizontal plane. atc: Semicylindrical cradle to which the dog is strapped. S1, S2: SUPports of the cradle with a built-in motor that can rotate the cradle about its cylinder axis through a desired angle. M: Magnet coil..xx _i R: Threaded rod holding the magnet coil. 0: "Gallows" for suspension of magnet coil. The wheels at the bottom of the gallows make it movable. percutaneously, in an external magnetic field. Descriptions of those parts of the apparatus and procedures which have been previously outlinedl12 will not be repeated. The first requirement is the placement of the flow probe at the desired location within the vascular system. This is made possible by fluoroscopy. X is the enclosure of the x-ray tube and I the image intensifier. The fluoroscopic image is viewed on a television screen not shown in Fig. 1. The "floating" table top, T of Fig. 1, makes it possible to move any desired portion of the animal over the x-ray tube (X) and the rotating cradle (C) can be used not only to insure a radiographically favorable orientation, but also to change the orientation of the flow probe with respect to the magnet to optimize the flow signal. The magnetic field is generated by a 452-turn coil (Al) wound of no. 11-gauge HID, polyester-insulated, aluminum wire (Magnet Wire Supply Co., Chatsworth, Calif.). The outer diameter of the windings is 40 cm and the inner diameter is 30 cm. The thickness of the windings is 5 cm. The coil is resonated at 400 Hz with a capacitance of 1.77 uf. A current of 0.9 amp is derived from a sinusoidal flow meter electronic channel as previously employed.1 The coil weighs 19 lb and is normally held in a "gallows" (G) as shown in Fig. 1. The magnet coil height is adjustable by suspension from a threaded rod (R) which is capped by a nut. The gallows-stand is mounted on a wheel base which makes it easily movable for adjustment of the coil position over the animal. Fig. 1 shows an unlabeled cardiac catheter emerging from the dog's left rear leg. The various sensors described below are introduced through it into the dog's aorta. The Flow Sensors. The principal experimental difficulty of the external magnetic field method of measuring blood flow is the transformer electromotive force (emf), which is induced at zero flow and, quite independently of the flow value, at any flow in the lead wires that convey the flow signal from the pick-up electrodes to the flow detector. It is subject to change as the relative position

3 VOL. 67, 1970 ARTERIAL BLOOD FLOW 1771 between the probe and the magnet is varied. This flow-independent emf is proportional to db/dt, and is thus in phase quadrature with respect to the sinusoidal magnetic field B. The flow signal, on the other hand, is in phase with the magnetic field B. This makes it possible to suppress the undesired flow-independent signal by means of a phase-sensitive detector so as to obtain a zero reading at zero flow. By placing a pick-up coil connected to the input of the phase-sensitive detector into the magnetic field, we can adjust the phase control of the detector until the output reads zero. This adjustment for rejection of the quadrature signal is at the same time the setting for optimal sensing of the flow signal. To facilitate the rejection of the zero-error voltage, it is essential to make the transformer emf as small as possible to begin with. The two different approaches to its rejection used at present have been described elsewhere.'.2 We shall thus confine ourselves to the illustration of the actual probes in Fig. 2 and to the description of a modification not previously described (Fig. 3). Fig. 2A shows a loop probe in which the error signal derived at zero flow from the electrodes E1, E2 would be very large if the lead wires were connected to E, and E2 in straightforward fashion. This signal is, however, reduced to zero by a special way of guiding one of the electrode leads.' The loop is made of a _..._ INCHES A X \ F: T R --I E~~~ E 2 E A B C D FIG. 2. Different types of intravascular flow sen ors used with anl external magnetic field. (A) Loop sensor. The loop collapses to pass through a narrow tube and expands in. the blood vessel to conitact its walls. El, E2 electrodes (7-mm long coils of mm platinum wire). (B) b-probe. El, E, electrodes. (C) L-probe emerging froii an intravascular catheter, C. Whei foilly inserted, the probe restumes the angulationi showin in B. To adjust it properly, as showii ill C, the probe stein is pulled back until the emerging portion assumes a 900 orientation relative to catheter C. This is accomplished with the aid of fluoroscopy. E,, E2 electrode,. (D) The original version of the bent wire probe, whose flow sensitivity is independent of angle of bend in dielectric conduits but not in arteries. El, E2 electrodes.

4 1772 PIrYSIOLOGY: KOLIN ET AL. PROC. N. A. S. LI S E L2 FIG. 3. Construction of the L-probe T1: Teflon shrink tubing shrunk over T2 "microthin" Teflon tubing. B: "Backbone" of 28-gauge hypodermic stainless steel tubing stiffening the tube T,. W: Twisted electrode lead wires. One of them emerges at L1 and is soldered to El, the other one emerges in the region of L2 and is soldered to E2. L1 and L2 are drops of solder to st1improve T, visibility during fluoroscopy. W B S: 0.2'5-mm stainless-steel stiffener. The probe is sealed by epoxy cement over the zones of L1 and L2. resilient steel frame (music wire 0.25 mm in diameter, encased in Teflon tubing 0.35-mm inner diameter, 0.65-mm outer diameter). E1, E2 are 7-mm long coils of mm platinum wire wound about the legs of the loop and platinized. The loop collapses as it passes through a narrow channel, such as a branch artery or a catheter, and resumes its original shape as it enters a wide channel, such as the aorta, where it expands to allow its electrodes to contact the artery wall at diametrically opposite points. This probe fits perfectly a wide range of artery diameters and can thus provide a measurement of the volume rate of flow since it records the induced voltage across the diameter of the blood vessel. Fig. 2D shows another sensor. It consists of a 28-gauge hypodermic tube encased in a "microthin" (Shamban Co.) teflon tube (0.35-mm inner diameter, 0.65-mm outer diameter) through the center of which a no. 40-gauge quad Tefloninsulated copper wire is passed. The tube and the wire form coaxial leads which convey the flow signal from the electrodes E,, E2 to the flow detector. Since there are no loops between the input leads, the transformer emf can be reduced to zero. When inserted into a tube, the electrodes E1 and E2 come to lie on diagonally opposite sides. If the magnetic field is perpendicular to the plane in which the end portion of this probe is bent, a flow signal of optimal intensity is recorded in a dielectric pipe. This signal is independent of the angle of the bend as long as the electrodes are located on opposite sides of the tube cross section, provided the uniform magnetic field extends far beyond the interspace between

5 VOL. 67, 1970 ARTERIAL BLOOD FLOW 1773 the electrodes. Under these coiiditioiis, the tube geometry and magiietic field distribution are substantially the same in any cross section along the pipe axis. It was believed that this independence of sensitivity from the angle of the bend applied to a conductive conduit as well as to a dielectric one.2 This turned out not to be the case. In fact, the flow signal in an artificial artery made of five layers of dialysis tubing of 64.0 ohm-cm resistivity, in an NaCl solution of 355 ohm-cm resistivity, is reduced to about 18% of the signal obtained in a dielectric pipe. Since this effect was not known earlier,2 it was not taken into account in affixing the calibration scale for the animal record (Fig. ic of ref. 2). To avoid this angular dependence effect, which will be considered in greater detail elsewhere, the probe angle must be maintained uniformly at 90. This yields flow signals in arteries of the same magnitude as obtained by loop probes of the type of Fig. 2A. The sensitivity drop reaches its minimum for this configuration and amounts to 20% in the case considered in Gessner's calculations for the flow of blood through an artery of infinite wall thickness, assuming the known resistivity for the blood and arterial wall tissues." -3 The requirement of a 900 bend becomes an insuperable obstacle for the design of Fig. ID, which is based on a sharply bent hypodermic tube. When such a tube, bent at right angles, is passed through a cardiac catheter of, say, no. 7 French size, the right angle is not restored when the probe end emerges. This difficulty was solved by changing the design and materials as depicted in Figs. 2B, 2C, and 3. The originally coaxial leads of the sensor have been replaced by a pair of twisted, teflon-insulated, 40-gauge copper wires, encased in Teflon tubing (Ti and T2), as shown in Fig. 3. The electrodes are 1-mm long coils of copper wire, 0.05 mm in diameter, or of platinum wire, mm in diameter. cemented with epoxy cement to the teflon tubing etched with "Tetra etch." The electrodes are platinized before use, as described.' One lead emerges at Li to be soldered to electrode El. The opening in the teflon tube is sealed with epoxy. The second lead proceeds toward the tip of the tube T2, to emerge at L2, bending upward to be soldered to electrode E2. Large drops of solder are used at Li and L2 to make these two points visible during fluoroscopy. A 0.25-mm music wire (S) serves to impart rigidity to the section between the electrodes El, E2. The tube is bent more than 900 as shown in Fig. 2B. When teflon is bent at a high temperature, it "remembers" its original shape and resumes it even after passing through a narrow catheter in which the bend is straightened out. Fig. 2C shows a probe, such as the one shown in Fig. 2B, emerging from an angiographic catheter of no. 7-gauge. By pulling the stem of the probe outward, one can regulate the angle of the bend and make it 900, as shown in Fig. 2C. L-shaped probes of the above description will function as volume-rate flow meters onlv if the electrodes are in contact with diametrically opposite walls of the blood vessel. The requirement of a precise fit makes this type of probe less desirable than the expanding-loop probe of Fig. 2A for volume flow measurements. They can, however, be used, without the requirement of fitting the artery, as velometers to measure the linear flow velocity, in cm/sec, in regions of demonstrated uniform velocity profile in pulsating arterial flow.4 5 The volume

6 1774 PHYSIOLOGY: KOLIN E'.AL. Panoc. N. A. S. rate of flow can then be computed from the measured flow velocity and the known diameter of the artery at the location of the flow sensor. In the experiments described in the following publication (Proc. Nat. Acad. Sci. USA, in press) the L-probes were used as velometers. Summary. Apparatus aind procedures for the measurement of the rate of blood-flow by means of sensors used in conjunction with an external magnetic field have been improved. A light magnetic coil is mounted adjustably over the animal. The magnetic field at the site of flow measurement can be determined by an intravascular magnetometer. The bent-wire probe has been modified to eliminate the uncertainty due to dependence of the sensitivity in the artery on the angle of the bend. It is found that, while the sensitivity is independent of the angle of the bend in dielectric conduits, it may fall to about 20%o of the value measured in a dielectric conduit in a conductive tube, such as an artery. The 90 angle of bend is optimal and easily reproducible and has been adopted. Abbreviation: emf, electromotive force. This work has been supported by Medical Testing Systems, Inc. 1 Kolin, A., Proc. Nat. Acad. Sci. USA, 65, 521 (1970). 2 Kolin, A., Proc. Nat. Acad. Sci. USA, 66, 53 (1970). Gessner, U., Biophys. J., 1, 627 (1961). 4McDonald, D. A., Blood Flow in Arteries (London: Edward Arnold Pub. Ltd., 1960). Kolin, A.,.1. 1). Archer, and G. Ross, Circ. Res., 21, 889, 1967.