Published in: Medical Imaging 1995: Physiology and Function from Multidimensional Images, Eric A. Hoffman, Editor, Proc. SPIE 2433, pages 303-308 (1995).




Cardiac Gated Ventilation

C.William Hanson, III* and Eric A. Hoffman**

*Department Anesthesia
Hospital of the University of Pennsylvania
3400 Spruce Street
Philadelphia, PA 1910

Division of Physiologic Imaging
Department of Radiology
University of Iowa College of Medicine
Iowa City, IA 52242

[ home | search | contact us ]

Table of Contents

ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
CONCLUSION
REFERENCES

ABSTRACT

There are several theoretic advantages to synchronizing positive pressure breaths with the cardiac cycle, including the potential for improving distribution of pulmonary and myocardial blood flow and enhancing cardiac output. We evaluated the effects of synchronizing respiration to the cardiac cycle using a programmable ventilator and electron beam CT (EBCT) scanning.The hearts of anesthetized dogs were imaged during cardiac gated respiration with a 50 msec scan aperture. Multi slice, short axis, dynamic image data sets spanning the apex to base of the left ventricle were evaluated to determine the volume of the left ventricular chamber at end-diastole and end-systole during apnea, systolic and diastolic cardiac gating. We observed an increased in cardiac output of up to 30% with inspiration gated to the systolic phase of the cardiac cycle in a non-failing model of the heart.

1. Introduction

Extrapolating from the mechanisms by which CPR generates and the intra-aortic balloon pump augments cardiac output, and from our early studies using the Dynamic Spatial Reconstructor showing that static lung inflation reduces the volume of the heart (all 4 chambers)1,2,3 we theorized that synchronizing positive pressure ventilation with the appropriate point in the cardiac cycle should enhance left heart function, and thus, perhaps, eliminate the need for a left ventricular assist device when a patient is on positive pressure ventilation. There is evidence that simultaneous chest compression and ventilation augments carotid blood flow in a cardiac arrest model, and previous studies by Pinsky et al. 4,5 have shown that cardiac cycle specific increases in intrathoracic pressure can augment cardiac output. However, since, in Pinsky's model, enhancement of cardiac output was found only in the failing heart, Pinsky hypothesized that the augmentation was due not to direct alterations in cardiac size but rather to increased myocardial blood flow associated with increased intrathoracic pressure.

The studies of Pinsky involved surgical invasion of the thorax, with potential changes in pleural pressures. Using advanced radiographic technology and a unique programmable ventilator, we were able to non-invasively study the effects of cardiac-gated ventilation on cardiac performance, ventilation- perfusion relationships and pulmonary blood flow. We have previous shown that computed tomography can be used to accurately6 evaluate the volume of the heart throughout the cardiac cycle1 and we used a similar approach to evaluate changes in stroke volume, ejection fraction and cardiac output in the present study via electron beam CT [EBCT]7 scanning and image processing via VIDA.8


2. Methods


2.1 Animal preparation

The animal use protocol was approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Six male mongrel dogs (weight 17+0,5' kg.) were studied, Anesthesia was induced with pentobarbital (25 mg/kg) and InnovarVet (0.2 mg/kg) and maintained with periodic injections on a regular schedule. During catheterization, the dogs were intubated and mechanically ventilated on room air with a Harvard constant volume ventilator. Extremity electrocardiographic leads (ECG) were placed on the right and left hind leg and the right fore leg. Carotid blood flow was measured with an ultrasonic flow probe (Infrasonics) placed around the carotid in the neck. A Millar catheter was advanced into the proximal pulmonary artery via the jugular vein for pressure monitoring. A second Millar catheter was placed via the carotid artery into the left ventricle, also for pressure monitoring. After the intravascular catheters were placed, an endotracheal tube was placed through a tracheal incision, with the tip lying above the carina. A urinary catheter was placed in the bladder and urine was collected in a bottle kept to the side of the scanner table. Fluids were administered through a femoral venous catheter such that volume loading kept up with urine flow, and arterial samples were drawn from the femoral artery after each ventilator mode change. Ventilation was adjusted to maintain a PCO2 between 35 and 45 torr. Pulmonary arterial and left ventricular pressures were monitored and recorded continuously during each scanning protocol along with a signal from the scanner showing the timing of scanning relative to physiologic events. At the time of catheterization, with the dog in a right lateral decubitus position, contrast media was injected into the left ventricle and a line was drawn on the dog's chest delineating the mitral valve plane.

2.2 Dye injection

Different catheters were positioned in preparation for various dye injection protocols: Cardiac wall motion: under fluoroscopic observation, a 7Fr side hole catheter was advanced via the external jugular into the superior vena cava. Coronary blood flow: a 7Fr side hole catheter was placed above the aortic valve. Pulmonary blood flow: a 7Fr catheter was placed in the proximal pulmonary artery. We will only report here on our findings related to augmentation of cardiac stroke volume.

2.3 Dynamic and Volumetric X-ray CT Imaging

Noninvasive evaluation of cardiac performance was performed with an electron beam computed tomography scanner (Imatron Corporation).5 This scanner operates an electromagnetically focused electron beam sweeping four parallel hemicylindrical (2100) tungsten targets, and a pair of solid state detector rings in a hemicylindrical configuration apposite the targets. One of the detector rings is double populated for high resolution scanning The machine operates with no target or detector motion, Configured to allow for the evaluation of regional pulmonary or cardiac blood flow and cardiac dynamics at multiple cross sections, the four tungsten targets are sequentially swept by the electron beam, with a beam sweep taking 50 msec and a reset interval between a target sweep taking 8 msec.

Cardiac Dynamics: Repetitive sweeps of the same target throughout a heart beat (scanning onset gated to the QRS complex of the ECG) yields two contiguous 8mm thick sections of the heart. With a 50msec target sweep, and an 8msec reset time, the heart in scanned at 10 points in the cardiac cycle spanning 572msec from end-diastole through systole and towards the subsequent end-diastole. Sweeping the next target ring repetitively over the subsequent cardiac cycle yields two more slices of the heart, scanned at 10 time points etc., until 8 levels of the heart have been acquired spanning 7.6cm of the apical to basal extent of the heart with 4mm thick gaps between pairs of contiguous slices. Success of such an approach is partially dependent upon the approximate stability of heart rate during scanning.

Myocardial Blood Flow: Instead of repetitively sweeping a single target throughout a cardiac cycle, all four target rings are swept in a single cardiac cycle with each target right swept in 50msec, reset time being 8msec, and then the next target swept in 50msec etc. until all four target rings are swept in 224msec. The first target sweep it timed to the QRS complex of the ECG and thus any pair of slices associated with a single target is always gathered at the same phase of the cardiac cycle. Timing of scanning and contrast bolus injection is such that the first sweep of the four target rings captures an image of the heart before any contrast arrives, and then subsequent sweeps captures the myocardial blush as contrast arrives according to the particular protocol being followed. In the case of myocardial blood flow evaluated by the protocol of Weiss,9 a slow continuous infusion of contrast in delivered over 8 seconds so that myocardial brightness rises to an equilibrium plateau.

2.4 Scanning and Experimental Protocol

Each dog was positioned on the scanner table within the gantry in a right side down, decubitus position, and oriented tangentially to permit short axis sectioning of the heart. The line drawn on the dog's thorax allowed the approximate positioning of the first slice to be scanned at the level and along the mitral valve plane. During slow drip contrast infusion, dogs were scanned and the position was adjusted such that a true short axis scan was obtained with a short axis acquisition judged by the circular appearance of the left ventricular cavity and the ability to obtain a basal section of the heart with minimal to no inclusion of atrial chamber area. Extremity ECG leads from the dog were connected to both the EBCT scanner and a Macintosh based ECG monitor system, Respiration was maintained by a computer programmable ventilator (CTP9000, CWE Inc., Ardmore, Pa.), developed in conjunction with our laboratory, designed specifically for physiologic gating. The ventilator consisted of a number of high frequency response needle valves under computer control. Through a set of macro-commands, the ventilator essentially any respiratory wave form could be programmed into the ventilator via its associated single board computer monitoring incoming physiologic signals and controlling the combination of needle valves opening and the opening and closing of a master valve leading to a continuous positive pressure air source. Standard volume ventilation was performed with tidal volumes of 15 cc's/ and adjusted to maintain a PaCO2 between 35 and 45 torr. Cardiac gated respiration was performed by initiating respiration in response to a signal generated when the electrocardiographic QRS complex exceeded a threshold value. During the experimental protocol, airway and intravascular pressures, ECG and EBCT scan-on data were recorded via a Macintosh IIfx computer running Labview (National Instruments). Prior to beginning the experimental protocol, a localization scan was obtained to determine scanning parameters including injection parameters for the heart cardiac dynamics of a given dog.

Studies were performed when the animal was: 1. apneic (APN); 2. with systolic-gated respiration (SGR: inspiration during systole and expiration during diastole); and 3. with diastolic gated respiration (DGR: inspiration during diastole and expiration during systole). Arterial blood gas values were obtained following each scan.

2.5 Image Analysis

Images were evaluated using image analysis software on the scanner console which allowed the user to trace the endocardial boarders of the heart and end-diastole and end-systole and reported chamber cross sectional areas. Tracing was done with the monitor brightness kept constant for all slice levels and all scanning protocols for a given dog. By multiplying the appropriate slice areas by the known slice thicknesses and spacing, cardiac chamber volume at end-diastole and end-systole was evaluated. In addition, images were transferred to a Sun computer running the VIDA software. In VIDA, the user draws a first guess at the location of the endocardial boarders. Then, the software smooths the trace and casts rays perpendicular to the trace such that along each ray, a point is found half way between the local brightness of the chamber center and the myocardial center. The local point along the original trace is adjusted or re-located to this "half-max" location. Through this process an objective endocardial boarder is found for each short axis section of the left ventricular chamber from base to apex at end-diastole and end-systole.

3. Results

Figure 1 shows a short axis section of the heart scanned during apnea (upper row), systolic gating (middle row), and diastolic gating (lower row). The dome of the diaphragm is traced at end-diastole (left column) and the trace is transferred to the end-systolic row (right). It can be seen that the diastolic diaphragm dome location at end-diastole and end-systole is identical for the apneic condition. It the middle row, one can see through the relative diaphragm dome position ad the diastolic trace location (middle right) that the diaphragm has moved caudally during cardiac systole indicating that inspiration occurred in systole. The opposite is seen in the bottom row. That is, the diaphragm has moved cephalad relative to the diastolic trace location as seen in the bottom right panel. Here, expiration is occurring during cardiac systole while inspiration occurred during cardiac diastole. The bright line in the left ventricular chamber represents the location of the pigtail left ventricular milar catheter tip manometer used to record LV pressure during the studies. These images also serve to demonstrate the accuracy with which we were able to identify the appropriate scanning orientation to acquire true short axis sections of the heart. It is the ability to truly section the heart along the short axis which contributed to our being able to accurately evaluated the volumes of the left ventricular chamber. To achieve this scanning plane, the dog was positioned in the scanner with the head to tail axis almost perpendicular to the longitudinal axis of the scanner table.

Figure 1: Short axis view of the left ventricle with location of the diaphragm dome identified at cardiac end-diastole and transferred onto the right column representing cardiac systole. The motion of the diaphragm dome in systole (right column) is shown by its change in position to the diaphragm dome represented by the superimposed trace. See text for further explanation. The three respiratory modes are: apnea (upper row), systolic gated respiration (middle row), and diastolic gated respiration (lower row).

Inspiration synchronized to cardiac ejection improved cardiac output, ejection fraction and stroke volume, without changing heart rate or cerebral blood flow. Cerebral blood flow appeared to be exquisitely autoregulated such that it could not be used as an index of the effect cardiac gated respiration was having on the heart. The improvement in cardiac output was apparent in both phases of respiratory gating, but was more prominent with Inspiration gated to occur during cardiac systole. When inspiration was timed to coincide with systole, stroke volume increased from 12.4+1.2 to 16.35+0.9 cc's/stroke, cardiac output from 693+80 to 908+75 cc's/min and ejection fraction from 52.3+2.2 to 66.1+ 2.38%.

Figure 2 shows the changes in stroke volume, cardiac output and ejection fraction with Inspiration gated to occur during cardiac systole. The asterisk indicates p<.05.

Myocardial blood flow was evaluated using a scanning and image analysis protocol technique described by Weiss et al.8 Preliminary data to date has shown that in changing from standard ventilation to respiratory gating in phase with ventricular systole, MBF increased from 84 to 195 ml/min/100gm (132%). A shift to diastolic gating resulted in a drop in MBF to 108 ml/min/100gm. In the same dog, a second change from standard ventilation to systolic gating caused an increase from 103 to 150 ml/min/100gm (50%). This is data from a single dog in which all scanning protocols for myocardial blood flow were successful.

4. Discussion

The results of this study are consistent with those of Pinsky, and indicate that inspiration gated to occur during systole can result in significant improvement in cardiac output, and possibly in myocardial blood flow. The mechanisms by which systolic gated respiration might enhance cardiac output are speculative, but improved cardiac emptying (systolic gating), improved cardiac filling (diastolic gating) or change in inotropic state (systolic gating) are all possibilities.The most likely explanation is improved emptying due to compression of the ventricles during the positive pressure portion of the respiratory cycle, which can be seen in figure two, which compares the cardiac and respiratory phase in a short axis section of the animal. The size of the ventricle is significantly small with SGR than with apnea during the same phase of respiration and cardiac contraction.

The advantages of using a noninvasive approach to the evaluation of cardiac and respiratory performance are the lack of interference with normal pleural and pericardial pressure and excursion. Using EBCT and computer based image analysis, we were able to closely approximate the conditions presented by the intact patient. We are not aware of any other method of cardiac evaluation to date which has the ability to volumetrically and dynamically evaluate the cardiac function in the manor demanded by this physiologic question. Dynamic, volumetric x-ray CT as provided by the Dynamic Spatial Reconstructor and the EBCT scanner offer a unique view of the intrathoracic milieu whereby the heart and lungs interact. Unlike the experiments of Pinsky, 4,5 we have shown that cardiac gated respiration improves cardiac output even in the non-failing heart.

5. Conclusion

The potential benefits of a ventilatory strategy which also improves cardiac output are self-evident. Mechanical ventilation is a well known cause of decreased cardiac output and hypotension in the critically ill patient, This approach to mechanical ventilation clearly deserves further investigation. Use of dynamic volumetric x-ray computed tomography allowed us to study the effects of cardiac gated respiration without the need to invade the thorax, and it is believed that our ability to study the heart-lung interactions in the undisturbed milieu allowed us to observe that which Pinsky's studies failed to detect. That is, even in the normal, non-failing heart, cardiac gated respiration improved cardiac output and stroke volume. It is presumed that with systolic gated respiration the augmentation of cardiac output comes from a direct compression of the heart by the expanding lungs and through the increase in intrapleural pressure associated with positive pressure lung inflation.

6. References

1. E. A. Hoffman and E. L. Ritman, "Invariant total heart volume in the intact thorax," Am J Physiol Vol. 249, pp. H883-890,1985.

2. E. A. Hoffman and E.L. Ritman, "Heart-lung interaction: Effect on regional lung air content and total heart volume," Annals of Biomedical Engineering, Vol. 15, pp. 241-257, 1987.

3. E. A. Hoffman, "Interactions: The integrated functioning of heart and lungs," In S. Sideman and R. Beyar (eds.) Interactive Phenomena in the Cardiac System p 347-364, Ch. 34 Plenum Press, New York, 1993.

4. M. R. Pinsky, G. M. Matuschak and M. Klain: "Determinants of cardiac augmentation by elevations in intrathoracic pressure," J Appl Physiol Vol. 58, pp.1189-1198,1985.

5. M. R. Pinsky, G. M. Matuschak, L. Bernardi and M. Klainr "Hemodynamic effect of cardiac cycle-specific increases in intrathoracic pressure," J Appl Physiol Vol. 6, pp.604-612,1986.

6. E. A. Hoffman and E. L. Ritman "Shape and dimensions of cardiac chambers via computed tomography: Role of image slice thickness and orientation," Radiology Vol. 155, pp.739-744,1985.

7. Boyd, D.P., and M.J. Lipton, "Cardiac computed tomography," Proceedings of the IEEE, Vol.71, pp.298-307, 1983.

8. Hoffman, E.A., D. Gnanaprakasam, K.B. Gupta, J.D. Hoford, S.D. Kugelmass, and R.S. Kulawiec, "VIDA: An environment for multidimensional image display and analysis," SPIE Proceedings, Vol. 1660, pp. 694-711, 1992.

9. Weiss, R. M., Otoadese, E.A., Noel, M.P., DeJong, S.C. and Heery, S.D.: "Quantitation of absolute regional myocardial perfusion using cine computed tomography." Vol. 23, pp.1186-93, 1994.





©1994-2000 Division of Physiologic Imaging, Dept. of Radiology, Univ. of Iowa


Papers | DPI Homepage | VIDA | NLM | Contact Us | Search



Last modified: Tue May 9 15:45:34 CDT