high frequency oscillators with adjustable waveforms

1 downloads 0 Views 1MB Size Report
adjustable wave-forms with high ratios of peak inspiratory to peak ... a simple sine-wave pressure curve. To produce a ... industrial vibration testing device used as a more powerful .... Chang HK, Weber ME, King M. Mucus transport by.
Br. J. Anaesth. (1989), 63, 38S-43S

HIGH FREQUENCY OSCILLATORS WITH ADJUSTABLE WAVEFORMS: PRACTICAL ASPECTS L. FREITAG, M. SCHROER AND J. BREMME At present high frequency oscillation is used mainly to improve gas exchange or to rest a lung to permit spontaneous ventilation to occur. Although some studies have been published [1-5], little is known of the effects of varying pressuretime characteristics and non-sinusoidalflowpatterns on cardiopulmonary function. However, another application of high frequency oscillation is dependent on the wave-form. It has been shown that high frequency oscillation with asymmetrical flow patterns (fig. 1) may improve mucus clearance even in the absence of natural clearance mechanisms such as ciliary beating [6—8]. In contrast, oscillations with inspiratory-biased flow profiles may propel sputum into the periphery of the lung. To move mucus cephalad by gas—liquid interaction, aflowpattern with higher expiratory than inspiratory peakflowsmust be obtained. It is the purpose of this paper to compare different types of ventilator in their ability to create adjustable wave-forms with high ratios of peak inspiratory to peak expiratory flow (PFI: PFE). We concentrate on the practical aspects, and confine this review to machines that are available commercially or can be built easily.

5000-

-5000-

-10000100

150

Time (ms)

FIG. 1. Example of periodic, non-harmonic wave form produced by linear motor. Expiratory biased flow pattern. PFE:PFI ratio = 2:1.

controlled over the cycle. This is virtually impossible with collector motors, because accelerating and braking consumes large quantities of energy, especially at higher frequencies. Stepper motors that could, theoretically, fulfil this task are not available in sizes required to drive pistons Mechanical devices large enough for ventilating the lungs of adult patients. Mathematically, any periodic motion For model studies a reciprocal pump with a quick-acting return has been used [6, 9]. We are can be separated into a series of superimposed not aware of any reports of the clinical application sine waves. Periodic, non-sinusoidal pressure of piston pumps different in principle from the curves may be created by combining two or more one described originally by Bohn and colleagues engines with different velocities. Figure 2 shows [10]. These machines use an eccentric motor with the concept of two independent, eccentric motors flywheel; the duty cycle (i: E ratio) is 1.0, creating with a common, adjustable shaft driving a piston. a simple sine-wave pressure curve. To produce a Motor speeds, crank radii, location of the pin biased flow profile with I:E ratios greater or less bearing on the connecting rod and piston volume than 1, the angular velocity of the motor must be are the variables which determine frequency, shape and amplitude of the periodic, non-harmonic wave-form. Although popular for research L. FREITAG, M.D.; M. SCHROER; J. BREMME, M.D.; Arzt fOr because of simplicity, the clinical usefulness of Lungen- und Bronchialhcilkunde, Ruhrlandklinik, Tflschcner crank based ventilators is limited, mainly because Wcg 40, D-4300 Essen 16, F.R.G. they are noisy and cannot be adjusted during use. Correspondence to L.F.

HIGH FREQUENCY OSCILLATORS

39S

V V CO , : 0J 2 • 1 2

FIG. 2. Double eccentric motor: modified piston pump with two independent motors with coupled crank mechanism. Non-sinusoidal pressure curves are created by varying motor speeds, crank radii and position of pin bearing.

Flow

Flow

wv

FIG. 3. Loudspeakers may be used as easily adjustable flow modulators with limited power.

Loudspeakers

Loudspeakers have been used frequently for basic physiological studies [11] and are incorporated in diagnostic tools for measuring respiratory impedance. It seems reasonable to use them as an adjustable high frequency ventilator because they can be controlled easily by standard poweramplifiers. We obtained the best results with two flat-membrane bass speakers (KEF B130) mounted face to face (figs. 3, 4). The frequency range covered is 5-200 Hz. PFI: PFE ratios of 1:4 with peak flows up to 7 litre s~l may be achieved and wave form modulation is easy. However, being an ideal diagnostic instrument, these devices are not suitable for ventilatory purposes. The main problems include limited power and the

extreme load-dependence of a loudspeaker. Even minor backpressures on the membrane, but especially PEEP and the patient's spontaneous ventilation, drive the magnetic coil out of its centre, making it impossible to maintain a stable wave-form. Electromagnetic flow generator

Figure 5 shows a modification of a loudspeakertype ventilator termed an electromagnetic flow generator (EFG, or "puffer"). An industrial vibration testing device moves a vaulted aluminium membrane inside a pressure chamber. The prototype shown in figure 4 is powerful enough to ventilate the lungs of patients at frequencies up to 20 Hz if lung damage is not too severe. For mucus mobilization, asymmetrical

40S

BRITISH JOURNAL OF ANAESTHESIA airway oscillations up to 30 Hz or artificial coughs can be generated with flow rates up to 15 litre s"1 using the "puffer" in parallel with an ordinary low-frequency respirator. The PFI :PFE ratio can be as large as 1:5, the highest of all devices tested. A pneumatic low pass filter diminishes the loadsensitivity of the device, but the wave form stability is far from satisfactory. When compared with loudspeakers (costing approximately 100 U.S. $) the device is not good value. An adequate magnet costs at least 4000 U.S. $ and an adequate power amplifier 1000 S. Furthermore, the device cannot be recommended for treatment of ARDS because it behaves as a pressure generator—that is, it is load-dependent and has limited power. Stiff lungs or obstructed lungs require larger flows and volumes than electromagnetic flow generators can deliver. Linear motor

FIG. 4. Prototypes of the ventilators discussed. 1 = Loudspeaker-type flow modulator. 2 = Linear motor with piston pump. 3 = Electromagnetic flow modulator. 4 = Digital Ventilator. 5 = Personal computer and power amplifier.

Low pass filter

FIG. 5. Electromagnetic high frequency flow generator: industrial vibration testing device used as a more powerful flow generator. An acoustic short-circuit reduces loaddependence.

The linear motor is also an electromagnetic device, but may be used to drive a piston (fig. 6). The advantage of this system is that the effect of an added load can be minimized by using an electronic feed-back to control motion. For experimental purposes, linear motors may be taken from the drives of old Winchester discs; these are robust, powerful and easy to adapt. The one that we used for our studies (surplus computer dealer, original manufacturer unknown) does require a large power supply (48 V, 20 A) and is very noisy. Controlling the volume displacement is not as simple as it may appear. The closed-loop cybernetic system has to be well engineered, because coupling of an electric load (solenoid) and a pneumatic load (impedance of tubing and patient's lung) with electronic feed-back circuitry easily results in self-oscillation and other alterations of the wave form. We found the frequency range is limited to 20 Hz with a peak-flow of 10 litre s"1 and a maximum stroke volume of 900 ml. The maximum PFI:PFE ratio is 2:5. Complex wave forms with high frequency components cannot be generated with the machine because the inertial mass of the mechanical parts becomes a limiting factor. Overall the linear motor is (if commercially available) an expensive but powerful and versatile solution. We have also used it to drive an external chest wall compression cuff. Digital Ventilator Completely different from the piston pump devices discussed above is the modified jet

HIGH FREQUENCY OSCILLATORS

41S Piston pump

Linear motor

Displacement

Actual point Set point Comparator amplifier

DAC Computer

FIG. 6. A linear motor may be used as a driver for a piston pump. The electronic feed-back circuitry is required to control the volume displacement, making it to some extent load-independent.

Pulse width modulation Valve 1 Compressed air



Tube adaptor with jet nozzles Compressed air



Tube

Flow T FIG. 7. Digital Ventilator working with two jet nozzles and pulse-width modulated solenoid valves.

ventilator that we had developed [12, 13] for patients in the ITU (fig. 7). An adaptor with two built-in jet-nozzles is attached to a tracheal tube or a mouth-piece. Their flows pass in opposite directions and they are controlled independently. One jet creates positive, the other one negative pressures in the airways. The jet streams are regulated by two fast-acting solenoid valves which are controlled by a microcomputer. When the inspiratory valve is activated, the machine functions as a simple positive pressure jet ventilator. However, the sequential opening of valves generates inspiratory and expiratory flows. With com-

puter controlled pulse width modulation any waveform may be created (fig. 8). The Digital Ventilator is remarkably versatile, as all wave form manipulations can be made by software. Capabilities are determined by the frequency response and maximum flow rates of the solenoid valves. In our prototype (fig. 4) we have used valves, provided by the Draeger company. We can achieve ventilating frequencies up to 45 Hz with peak flow rates up to 5 litre s~!. The performance can probably be enhanced if several solenoid valves are used in parallel and the Venturi adaptor is optimized. However, although

42S

BRITISH JOURNAL OF ANAESTHESIA

Wffltf, frtt HFJV

HFO

Vibration

Pseudo-CMV

FIG. 8. Strip chart recording (pressure v. time in test lung) of the Digital Ventilator. A variety of different flow curves can be accomplished by software changes.

it is suitable for ventilating the lungs of a patient, the device has limited ability to produce the complex wave forms required for mobilization of mucus. With our prototype, the necessary asymmetry of oscillation (much higher expiratory than inspiratory flow) can scarcely be maintained if the patient is breathing spontaneously. Fluctuations in airway pressure disturb the Venturi effect on which the wave form stability depends. Very high frequency components cannot be created because the inertia of the gas is too high. The load dependence of the Digital Ventilator is tolerable for most clinical applications, but we have not been able to ventilate lungs adequately when the compliance was less than 200 ml kPa"1. SUMMARY

We have shown that several types of high frequency oscillator can be modified to produce non-sinusoidal flow patterns. We are convinced that the ability to adjust the wave form is a useful feature of a high frequency ventilator. As the required energy is related to (frequency)2 all the machines we have examined are large, heavy and noisy. Some problems could probably be overcome by engineering skill. For this reason we have not presented detailed power charts of our machines (built mostly by us). However, one has to admit that none of the devices is sophisticated enough to satisfy a physician who is not enthusiastic for high frequency ventilation. On the other hand, we have treated several patients suffering from bronchiectasis using a linear motor piston heavier than the patient. Asymmetric oscillations helped to clear the airways of purulent sputum. Despite all the noise and unsophisticated appearance of the machine, it was accepted well

by the patients because they felt relief. This suggests that, regardless of all obstacles, research in HFV technology should be continued. ACKNOWLEDGEMENTS We wish to thank John Lehr (Harvard Medical School), from whom we borrowed the linear motor that he had developed, and Volker Freiburg (University of Bochum) for excellent technical assistance. The studies were supported by the Deutsche Forschungsgemcinschaft, the E. Jaeger company Wurzburg and the Arbeitsgemeinschaft zur Foderung der Pneumologie an der Ruhrlandklinik.

REFERENCES 1. Slutsky AS, Kamm RD, Rossing TH. Effects of frequency, tidal volume and lung volume on CO, elimination in dogs by high frequency (2-30 Hz), low tidal volume ventilation. Journal of Clinical Investigation 1981; 68: 1475-1484. 2. Perry, M, Blue P, Kindig Ghaed N. Pressure wave form correlation with xenon washout time in a physical model of high frequency oscillation. In: Proceedings of an International Symposium on High Frequency Ventilation. New York City, 1983; 127-128. 3. Spoelstra AJG, Tamsa TJA. HFJV: the influence of gas flow, inspiration time and ventilatory frequency on gas transport in healthy and anaesthetized dogs. British Journal of Anaesthesia 1987; 59: 1298-1308. 4. Paloski WH, Slosberg RB, Kamm RD. Effects of gas properties and waveform asymmetry on gas transport in a branching tube network. Journal of Applied Physiology 1987; 62: 892-901. 5. Vcnegas JG, Yamada Y, Custer J, Hales CA. Effects of respiratory variables on regional gas transport during high-frequency ventilation. Journal of Applied Physiology 1988; 64: 2108-2118. 6. Chang HK, Weber ME, King M. Mucus transport by high frequency nonsymmetrical oscillatory airflow. Journal of Applied Physiology 1988; 65: 1203-1209. 7. Freitag L, Kim CS, Long WM, Venegas JG, Wanner A. Mobilisation of mucus by airway oscillations. Ada Anaesthesiologica Scandinavica 1989; 90: 93-101.

HIGH FREQUENCY OSCILLATORS 8. Freitag L, Long WM, Kim CS, Wanner A. Removal of bronchial secretions by asymmetric oscillations at the airway opening. Journal of Applied Physiology 1989; 67: 614-619. 9. Mablie H, Ocvirk FW. Mechanisms and Dynamics of Machinery, 3rd Edn. New York: Wiley, 1987. 10. Bonn DJ, Miyasaka K, Marchak BE, Thompson WK, Froese AB, Bryan AC. Ventilation by high-frequency oscillation. Journal of Applied Physiology 1980; 48: 710-716.

43S 11. Landser FJ, Nagels N, Demedts M, Billiet L, Van de Woestijne KP. A new method to determine frequency characteristics of the respiratory system. Journal of Applied Physiology 1976; 28: 289-301. 12. Freitag L, Wcndt M, Dankwart F. Digital ventilation. In: Lawin P, ed. Maschinelle Beatmung gestern-heute-morgen. Thieme INA 48, 1984; 327-334. 13. Wendt M, Freitag L, Dankwart F. Digital-ventilation. In: New Perspectives in High Frequency Ventilation. Rotterdam: M. Nijhoff Publ., 1982; 172-177.

Suggest Documents