| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of 1Clinical Physiology, 2Radiology, and 3Radiophysics, Lund University Hospital, Lund SE-22185; and 4Inovacor, Stockholm SE-18592, Sweden
Submitted 25 November 2003 ; accepted in final form 5 March 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
magnetic resonance imaging; left ventricle; output
In 1932, Hamilton and Rompf (11) described a relative constancy of the total heart volume during the cardiac cycle in frogs, turtles, and dogs. They concluded that through the motion of the atrioventricular (AV) plane, the heart was able to pump blood but still maintain the same volume. Subsequent noninvasive investigations in cats (10) and dogs (12, 14) were concordant with this initial finding of a relatively consistent cardiac volume. Investigations in humans with the use of computed tomography (15) and MRI in ventilated patients (18) have suggested that a volume variation of 8–13% may occur between diastole and systole. However, recent work (7) with the use of high-resolution MRI found a lower volume variation of 5% with the largest variation calculated to be at the left side using individual chamber volume measurements at end systole and end diastole (7).
These studies confirm the presence, but an unclear extent and timing, of total heart volume variation throughout the cardiac cycle at rest. Additional physiological information may be obtained using measurements of blood flow that are easily obtained using MRI.
Therefore, the goals of this study were to investigate the magnitude, timing, and contributors of the total heart volume change during the cardiac cycle in healthy humans with the use of both the MRI-planimetric method as in earlier studies as well as quantitative velocity mapping MRI (21). This will allow the opportunity to explore total heart volume variation from a structural (planimetry) and functional (flow) perspective. The use of both of these principally different techniques in the same subjects gives the possibility to identify whether the volume change is due to a physiological phenomena or a measurement error inherent to the planimetric method (which could explain the large variation between studies).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Eight healthy volunteers (26–47 yr old, 5 men and 3 women) were examined after approval of the local ethics committee. Subjects underwent cardiac MRI in the supine position. Short-axis gradient-echo cine imaging encompassing the entire heart from the base of the atria to the apex of the ventricles was followed by blood flow measurements using velocity mapping in all vessels leading into and out of the heart. Changes in total heart volume throughout the cardiac cycle were calculated directly by planimetry of the gradient-echo cine images and indirectly by subtracting all of the blood flow out of the heart (aorta and pulmonary artery) from all blood flow into the heart (inferior and superior vena cava and the lung veins).
Magnetic Resonance Imaging
A 1.5-T MRI scanner (Magnetom Vision; Siemens, Erlangen, Germany) with 25 mT/m maximum gradient strength and 600-µs gradient ramp time was used to acquire all images.
Volumetric imaging.
ECG-triggered gradient-echo sequences were used to obtain cine images during end-expiratory apnea (
15 s). The number of phases was determined by the R-R interval, resulting in 15–21 images per heart cycle. The typical imaging parameters were the following: 100-ms repetition time (echo sharing resulting in phases every 50 ms), 4.8-ms echo time, 30° flip angle, and 10-mm slice thickness (edge to edge). Short-axis images (Fig. 1, C and D) were taken from cardiac base to ventricular apex perpendicular to the left ventricular longitudinal axis (Fig. 1, A and B). Three long-axis images were obtained in the two-chamber, four-chamber, and the left ventricular aortic-outflow-tract views.
|
Figure 2 shows an example of the velocity mapping technique in an image plane perpendicular to the inferior caval vein. The grayscale values of each pixel in these images (Fig. 2B) are directly proportional to velocity. Thus within a region of interest (ROI), the average blood flow velocity for any given time point throughout the cardiac cycle can be calculated. Knowledge of the measured area will allow calculation of absolute blood flow (average velocity x area = average flow) (Fig. 2C). Two ROIs for background correction of field inhomogenities were drawn in stationary tissue at equal length from the vessel investigated, as previously described (2). The average signal intensity of these ROIs in nonmoving tissue were set to represent 0 cm/s.
|
Flow measurements. Two different evaluation platforms were used (RADGOP; Context Vision, Linköping, Sweden, and Scion Image). Each blood vessel was manually outlined in the anatomic images for each time point with flow calculation undertaken in the corresponding velocity-mapped images. To determine the change in heart volume, the difference between the sums of blood flow into the heart (pulmonary and caval veins) and out of the heart (aorta and pulmonary artery flow) was calculated over the cardiac cycle. Total heart volume change by the flow method divided by total heart volume at end diastole by the volumetric method gives the percentage of total heart volume change for the flow method.
Isolated contribution of systole to change in total heart volume. To calculate the total heart volume change that would occur if there were no cardiac inflows to the atria during systole, the combined volume ejected into the aorta and pulmonary artery in systole (the ventricular stroke volumes obtained by the flow method) was divided by the total heart volume at end diastole obtained by the volumetric method. To calculate how large the filling of the heart is during systole, the combined inflows to the heart were divided by the total heart volume at end diastole by the volumetric method. To calculate the percentage of the stroke volumes that is secured by the heart by filling of the atria during ventricular systole, the combined inflows during systole were divided by the stroke volumes obtained by the flow method.
Chamber contribution to change in total heart volume. Flow measurements were used to calculate the contribution to the heart volume change from the left and the right side of the heart. The net volume difference between inflow and outflow to the left side of the heart (pulmonary veins and aorta) from end diastole to end systole was divided by the total cardiac volume difference from end diastole to end systole. This provided the contribution of the left side (atrium and ventricle) to the total heart volume change in percent.
Validation of short-axis measurements and interobserver variability. To validate short-axis measurements [which have proven to be the best plane to determine left ventricular volumes (9, 19) and therefore are commonly used in cardiac MRI], one subject was examined using the frontal, axial, sagittal (5-mm image slice thickness), as well as the short-axis imaging planes (10-mm image slice thickness). The variation in total heart volume was compared between each of these imaging planes. Interobserver variability was performed by two independent observers in all subjects for volumetric measurements and in the first four subjects for flow measurements.
Statistical Analysis
Continuous variables are presented as means ± SE and with the range of the variables. The interobserver or method concordance was calculated using the intraclass correlation coefficient (ICC) (25). Bland-Altman analysis (5) and ICC were used to test whether the two methods used to measure changes in total heart volume (volumetric and flow) differed. Paired t-test was used to test whether the changes in apex base length were significant. Pearson's correlation was used to examine the relationship between relative contribution to total heart volume change and starting slice volume.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Good agreement (ICC 0.91) between imaging planes was observed for total heart volume and its variation and was consistent with previous investigations of interstudy variability (4, 24). Therefore, only one imaging plane (short axis) was used for subsequent subjects. The interobserver variability of planimetry measurements of total heart volume using short-axis images was good (ICC 0.98) and consistent with earlier studies of interobserver agreement of MRI measurement (4, 6, 24). The interobserver variability of flow measurements, ICC >0.99, was also acceptable. The agreement between observers is visually demonstrated by the original data in Fig. 3 where all in- and outflows measured by two observers in one subject is shown.
|
The overall total heart volume variation during a cardiac cycle did not differ between the volumetric and the flow methods: 8.2 ± 0.8% (range 4.8–10.6%) versus 8.8 ± 1.0% (range 5.6–11.8%) (Table 1). ICC was 0.999 and the difference according to Bland-Altman analysis was –0.6 ± 1.0% (SD). The magnitude and time course of the total heart volume variation were essentially similar in all subjects (Fig. 4). It should be noted that the investigated subjects were at rest, with cardiac output of 6.4 ± 0.3 l/min, range 4.7–7.4 l/min; cardiac index 3.4 ± 0.1 l·min–1·m–2, range 2.9–3.7 l·min–1·m–2 and low heart rates, range 53–71 beats/min.
|
|
|
The proportion of total heart volume variation was greater on the left side of the heart, 61 ± 2% (range 52–70%) compared with the right (Table 1). The predominant volume changes were visually observed where the heart was most adjacent to the lungs, whereas the regional changes in volume at the borders to the liver were less and were essentially absent near the thoracic wall (Fig. 8B).
|
The largest volume change during systole was located where the diameter of the heart was largest, i.e., at the base of the ventricles. This coincided with the region of the AV plane movement (Figs. 1 and 6). The contribution to change in total heart volume from each individual short-axis slice was relative to the size of that slice (r2 = 0.43) (Fig. 7) resulting in a proportional decrease in volume in all parts of the heart in systole.
|
|
Longitudinal Shortening of the Heart
There was only a minor change in the length of the heart, 0.9 ± 0.5% (range –2 to 3%; P = 0.07) from apex to base, and the apex of the ventricles was essentially stationary (Table 1).
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
8% (range 5–11%) in total heart volume between diastole and systole in healthy individuals measured noninvasively using cine-MRI. Importantly, both the planimetric and flow-based methods independently show similar results for total heart volume variation in each individual subject and the reproducibility of each of these techniques was shown to be high. The variation in total heart volume change between subjects likely demonstrates a true physiological variation because of the concordant findings from two independent methods. The variation arises from a decrease in total heart volume in systole due to a discrepancy in blood flow into and out of the heart with no significant longitudinal shortening of the heart during this process. The major contributor to the volume change is the region around the AV plane movement with a leftsided predominance. For any given cardiac MRI short-axis imaging plane, the change in volume was essentially proportional to the starting volume in that plane. Earlier Studies of Total Heart Volume
Since the days of Da Vinci (27), researchers have attempted to explain the fundamental properties of cardiac function. Much of this work has concentrated on the mechanics of individual chambers; however, the properties of total heart volume and shape are less well explored.
Earlier studies undertaken in this area using animal and human subjects have produced conflicting results regarding the variation of cardiac volume. Some reported no variation (11), some reported a variation within the measurement error (14), and others suggested that a variation indeed exists, although the total heart volume variation varies between studies from 5 to 13% (7, 12, 15, 18). In 1932, Hamilton and Rompf (11) did a study of the heart in frogs, turtles, and dogs and concluded that the variation of total heart volume during the heart cycle was minimal and that much of the cardiac output was generated by movement of the AV plane. Subsequent work by Gauer (10) expanded these initial findings by acquiring fluoroscopic ventriculograms of cat hearts and concluded that the "outside dimensions of the heart...identical" between systole and diastole. Further studies by Hoffman et al. (12, 15) in dogs were consistent with this observation with a total heart volume variation of 5% that increased with atrial fibrillation. However, Leithner et al. (18) demonstrated in humans a variation of up to 13% of intracavitary volumes but suggested that this could not be considered as total heart volume and therefore was guarded in drawing any conclusions about true total heart volume variation. Bowman et al. (7) recently proposed that the total chamber volumes in the pericardium may be less constant than the blood pool content. In the present study, however, we did not find any significant difference between total heart volume variation obtained by flow or planimetric measurements.
The variation between subjects in total heart volume change is likely a true physiological finding because two principally different and independent techniques (flow and planimetry) yielded highly concordant results. This intersubject variation may explain the differing results in total heart volume change found in previous studies.
Physiological Relevance
A low total heart volume variation will lower the energy required to move surrounding tissues during filling and emptying of the heart (12). Our qualitative visual assessment and quantitative findings (Table 1) have shown that the largest volume variation occurred in the region of the AV plane and slightly more on the left-hand side and is consistent with previously published findings (13). At higher heart rates and cardiac output, the momentum of the blood created by the descent of the AV plane can be used for a more rapid filling of the expanding atria and the reverse motion of the AV plane. Thus at higher heart rates, blood will have a higher momentum, decreasing the requirement of outer volume change in the surrounding tissue, which would save energy. In addition, higher heart rates may result in a lower total heart volume change because of a shortened diastolic phase, and this possibly could make the outflow less pulsatile, and hence, synchronize the in- and outflow (20). This could result in a lesser total heart volume change at higher frequencies. This is in line with the findings of Brecher (8), Nilsson et al. (22), and the conclusions of Gauer (10) that faster and smaller hearts secure 80% of their stroke volume during ventricular systole. In the present study in humans (larger species and therefore larger hearts and slower heart rates) 64% (range 55–73%) of the stroke volumes from both chambers were secured into the atria during systole (reservoir function of both atria). The reservoir function of the left atrium in percentage of the stroke volume of the left ventricle has previously been reported to be 38% as assessed by Doppler echocardiography (23) and 41% as assessed by MRI (17). Thus the reservoir function of the right atrium is higher than the left atrium. This can be visualized in the flow curves as a more bipolar flow in the caval veins with the highest peak in systole and a more continuous flow in the pulmonary veins.
In Doppler examination of caval and pulmonary veins, a similar biphasic pattern of blood flow during systole and diastole into the heart as in the present study has been found (3) and also alteration of cardiac filling according to heart rate (1). At an increased heart rate, proportionally greater systolic atrial filling over diastolic flow occurs compared with that which occurs at lower heart rates. These findings with echocardiography are consistent with the previously stated relationship between cardiac volume variation and heart rate, as increased systolic atrial filling would minimize volume variation.
Further Studies
Studies with MRI under dobutamine and/or atropine stress would give the opportunity to tease out the effect on total heart volume variation of higher heart rates and cardiac outputs discussed above. This could give new information on the cardiac energetics in man under different physiological conditions.
Limitations
The volumetric images in this study were acquired during end-expiratory apnea. This potentially may affect cardiac output, pulse rate, and introduce a potential "valsalva-like" effect on the filling of the cardiac chambers. In addition, the temporal resolution of the volumetric image sequences was 50 ms and thus may be an undersample of the true end-diastolic and end-systolic dimensions of the heart. Despite these potential errors, the flow-based measurements that have a higher temporal resolution (30–40 frames per heart cycle) and are acquired during free breathing were in close concordance with the volumetric measurements and therefore suggests that these errors are likely to be small at most. It should be noted that the volumetric images are acquired during 15–19 heartbeats per slice and that the flow images are acquired during 256 heartbeats (or 512 when the images were obtained over two R-R intervals). Other imaging limitations inherent to cardiac MRI were occasionally present during this study, including the "partial volume effect," which results from a composite image of 1-cm imaging planes. This effect, however, is also likely to be small as demonstrated by the validation of the 1-cm short-axis slice thickness with 0.5-cm slice thickness measurements in three orthogonal planes.
In conclusion, the present study has demonstrated in humans that total cardiac volume, during a full cardiac cycle at low cardiac output and heart rate, diminishes 8% (range 5–11%) during systole and predominantly occurs in the region of the AV plane and on the left side. The total heart volume variation may relate to the efficiency of energy use by the heart to minimize displacement of surrounding tissues while accounting for the energy required to draw blood into the atria during ventricular systole.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. L. Ritman, A. Pasipoularides, T. Arts, T. Delhaas, P. P. Sengupta, B. K. Khandheria, A. J. Tajik, A. Boussuges, and J. Regnard To the editor: functional imaging(FI) combines imaging datasets and computational fluid dynamics to simulate cardiac flows. J Appl Physiol, September 1, 2008; 105(3): 1015 - 1015. [Full Text] [PDF]
|
|
|
|
 |
|
 B G. Svealv, E L Olofsson, and B Andersson Ventricular long-axis function is of major importance for long-term survival in patients with heart failure Heart, March 1, 2008; 94(3): 284 - 289. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Psychidis-Papakyritsis, A. de Roos, and L. J. M. Kroft Functional MRI of Congenital Absence of the Pericardium Am. J. Roentgenol., December 1, 2007; 189(6): W312 - W314. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Carlsson, M. Ugander, E. Heiberg, and H. Arheden The quantitative relationship between longitudinal and radial function in left, right, and total heart pumping in humans Am J Physiol Heart Circ Physiol, July 1, 2007; 293(1): H636 - H644. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Carlsson, M. Ugander, H. Mosen, T. Buhre, and H. Arheden Atrioventricular plane displacement is the major contributor to left ventricular pumping in healthy adults, athletes, and patients with dilated cardiomyopathy Am J Physiol Heart Circ Physiol, March 1, 2007; 292(3): H1452 - H1459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Carlhall, L. Wigstrom, E. Heiberg, M. Karlsson, A. F. Bolger, and E. Nylander REPLY Am J Physiol Heart Circ Physiol, November 1, 2006; 291(5): H2551 - H2552. [Full Text] [PDF]
|
|
|
|
|
|
B. Wandt, L. A. Brodin, and S. Lundback Misinterpretation About the Contribution of the Left Ventricular Long-Axis Shortening to the Stroke Volume Am J Physiol Heart Circ Physiol, November 1, 2006; 291(5): H2550 - H2550. [Full Text] [PDF]
|
|
|
|
|
|
M. M. Riordan and S. J. Kovacs Relationship of pulmonary vein flow to left ventricular short-axis epicardial displacement in diastole: model-based prediction with in vivo validation Am J Physiol Heart Circ Physiol, September 1, 2006; 291(3): H1210 - H1215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. A. Waters, A. W. Bowman, and S. J. Kovacs MRI-determined left ventricular "crescent effect": a consequence of the slight deviation of contents of the pericardial sack from the constant-volume state Am J Physiol Heart Circ Physiol, February 1, 2005; 288(2): H848 - H853. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. van Dam and A. van Oosterom Volume conductor effects involved in the genesis of the P wave Europace, January 1, 2005; 7(s2): S30 - S38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. W. Bowman and S. J. Kovacs Prediction and assessment of the time-varying effective pulmonary vein area via cardiac MRI and Doppler echocardiography Am J Physiol Heart Circ Physiol, January 1, 2005; 288(1): H280 - H286. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
). This was done by mapping inflow and outflow (A) and calculating the cumulative volume (B) as the sum of inflow and outflow at each phase of the cardiac cycle. A: total flow leaving the heart (the sum of the flows of the aorta and the pulmonary artery) is pulsatile and the total flow entering the heart (the sum of the flows of the cava and pulmonary veins) is more continuous. Positive values are inflow and negative values outflow. B: there is a discrepancy in time of the cumulative values of the volume entering and leaving the heart due to the difference in flow patterns described in A. The cumulative curves diverge during systole and converge during diastole. This results in a decrease in total volume in systole that returns to the presystolic volume during diastole. C: volume difference (
) is negative during systole, meaning that total heart volume (
