* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Quantification of left and right atrial kinetic energy using four
Survey
Document related concepts
Coronary artery disease wikipedia , lookup
Cardiac contractility modulation wikipedia , lookup
Heart failure wikipedia , lookup
Electrocardiography wikipedia , lookup
Cardiac surgery wikipedia , lookup
Myocardial infarction wikipedia , lookup
Mitral insufficiency wikipedia , lookup
Hypertrophic cardiomyopathy wikipedia , lookup
Quantium Medical Cardiac Output wikipedia , lookup
Lutembacher's syndrome wikipedia , lookup
Atrial septal defect wikipedia , lookup
Ventricular fibrillation wikipedia , lookup
Dextro-Transposition of the great arteries wikipedia , lookup
Atrial fibrillation wikipedia , lookup
Arrhythmogenic right ventricular dysplasia wikipedia , lookup
Transcript
J Appl Physiol 114: 1472–1481, 2013. First published March 14, 2013; doi:10.1152/japplphysiol.00932.2012. Quantification of left and right atrial kinetic energy using four-dimensional intracardiac magnetic resonance imaging flow measurements Per M. Arvidsson,1 Johannes Töger,2 Einar Heiberg,2 Marcus Carlsson,1 and Håkan Arheden1 1 Department of Clinical Physiology, Lund University, Skane University Hospital, Lund, Sweden; and 2Department of Numerical Analysis, Centre for Mathematical Sciences, Lund University, Lund, Sweden Submitted 30 July 2012; accepted in final form 6 March 2013 four-dimensional phase-contrast magnetic resonance; cardiac magnetic resonance; energy; cardiac function ON A CONCEPTUAL LEVEL, the heart aims to deliver a pressurized amount of blood to the systemic and pulmonary circulation. The external work performed by the heart can be divided into kinetic energy (KE) and stroke work. Stroke work constitutes the vast majority (about 99%) of the external work of the left ventricle (LV) at rest (30, 31) and slightly less (94%) in the right ventricle [RV (6, 31)]. KE is a measure of how much potential energy is used to accelerate a volume of blood, for example, the blood that enters the ventricles from the atria during ventricular diastole. In other words, KE is a measure of the amount of work that is directly involved in moving blood, due to its inertia. KE thus relates to the momentum of blood, which explains why blood continues to flow out of the heart in late ventricular systole (34) and has been postulated as an important factor for cardiac pumping efficiency (23), espe- Address for reprint requests and other correspondence: H. Arheden, Dept. of Clinical Physiology, Skane Univ. Hospital, Lund Univ., 22185 Lund, Sweden (e-mail: [email protected]). 1472 cially so during exercise (22). Atrial KE constitutes a previously unaccounted part of the ventricular external work because ventricular contraction drives atrial filling (2, 36). The importance of the conservation of KE in the looped heart has recently been under debate (24, 39, 40). Recent studies have demonstrated rotational flow patterns in atrial blood and suggested that this arrangement may be beneficial for energy conservation (14, 23). However, the energy involved in rotational flow has not yet been quantified. We hypothesized that the organization of atrial blood into rotational flow structures may function as a hydraulic flywheel. Such a mechanism could conserve blood KE, which may be increasingly important for ventricular filling during exercise. Normal physiology aside, KE and pressure are independent parts of the external work of the heart, wherefore KE may be affected in disease with normal blood pressure. Interestingly, while cardiac stroke work has been thoroughly investigated and holds an important place in the clinic, intracardiac KE remains largely unexplored. The current debate on the potential value of energy conservation lacks data (23, 39). Therefore, the importance of KE in cardiac function remains unclear. Cardiac magnetic resonance (CMR) with three-dimensional, time-resolved phase-contrast sequences (4D PC) is the only imaging modality that allows for direct quantification of KE in a volume of blood (14, 25). Recent studies quantified the KE of ventricular blood during ventricular diastole (12, 13) and over the entire cardiac cycle (6) and provided new insights into cardiac energetics and ventricular function. Likewise, studies of atrial KE may provide new insight into the mechanisms of atrioventricular interaction. Therefore, the aims of this study were to quantify the KE of the left and right atria over the entire cardiac cycle using four-dimensional phase-contrast magnetic resonance imaging (4D PC-MRI), to measure the fraction of KE involved in rotational flow, and to identify the mechanisms contributing to atrial KE. METHODS Study population. Fifteen healthy volunteers (age 23–52, 7 females) were included in the study. Subject characteristics are summarized in Table 1. All subjects had blood pressure ⱕ140/90 mmHg, normal electrocardiogram (ECG), were without medication, and had no history of cardiovascular or systemic disease. The local institutional committee approved the study, and written informed consent was obtained from all subjects, in accordance with the Helsinki declaration. Magnetic resonance imaging. The subjects underwent CMR examination at rest in the supine position in a 3 T Philips Achieva scanner. We acquired 4D PC flow images from a box covering the entire heart and balanced steady-state free precession cine images in the twochamber, three-chamber, four-chamber, and short-axis views. The sequences used have been previously described in greater detail (6). 8750-7587/13 Copyright © 2013 the American Physiological Society http://www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 Arvidsson PM, Töger J, Heiberg E, Carlsson M, Arheden H. Quantification of left and right atrial kinetic energy using fourdimensional intracardiac magnetic resonance imaging flow measurements. J Appl Physiol 114: 1472–1481, 2013. First published March 14, 2013; doi:10.1152/japplphysiol.00932.2012.—Kinetic energy (KE) of atrial blood has been postulated as a possible contributor to ventricular filling. Therefore, we aimed to quantify the left (LA) and right (RA) atrial blood KE using cardiac magnetic resonance (CMR). Fifteen healthy volunteers underwent CMR at 3 T, including a four-dimensional phase-contrast flow sequence. Mean LA KE was lower than RA KE (1.1 ⫾ 0.1 vs. 1.7 ⫾ 0.1 mJ, P ⬍ 0.01). Three KE peaks were seen in both atria: one in ventricular systole, one during early ventricular diastole, and one during atrial contraction. The systolic LA peak was significantly smaller than the RA peak (P ⬍ 0.001), and the early diastolic LA peak was larger than the RA peak (P ⬍ 0.05). Rotational flow contained 46 ⫾ 7% of total KE and conserved energy better than nonrotational flow did. The KE increase in early diastole was higher in the LA (P ⬍ 0.001). Systolic KE correlated with the combination of atrial volume and systolic velocity of the atrioventricular plane displacement (r2 ⫽ 0.57 for LA and r2 ⫽ 0.64 for RA). Early diastolic KE of the LA correlated with left ventricle (LV) mass (r2 ⫽ 0.28), however, no such correlation was found in the right heart. This suggests that LA KE increases during early ventricular diastole due to LV elastic recoil, indicating that LV filling is dependent on diastolic suction. Right ventricle (RV) relaxation does not seem to contribute to atrial KE. Instead, RA KE generated during ventricular systole may be conserved in a hydraulic “flywheel” and transferred to the RV through helical flow, which may contribute to RV filling. Atrial Kinetic Energy in Humans Table 1. Subject characteristics Subject No. Gender Age, yr Heart Rate, beats/min BSA, m2 Blood Pressure, mmHg 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 F F M M M M M F M M M F F F F 52 36 24 23 29 28 28 25 26 25 25 25 30 26 28 58 60 57 58 78 60 69 69 72 77 61 58 70 85 69 1.6 1.9 2.1 2.2 2.1 1.8 1.8 1.8 2.3 2.1 2.2 1.9 1.8 1.8 1.6 140/90 120/65 120/65 110/70 130/80 100/70 120/90 130/90 140/75 125/75 120/70 95/65 120/70 110/70 105/60 M, male; F, female; BSA, body surface area. Arvidsson PM et al. 1473 We hypothesized that atrial KE during ventricular systole would be dependent on atrioventricular (AV) plane velocity and atrial volume. Atrial KE during ventricular diastole was hypothesized to vary with ventricular wall mass and systolic atrial KE. To determine the contribution of AV plane velocity to atrial KE, the AV plane displacement (AVPD) was measured in all subjects. The measurements were performed in the three long-axis images, using the same method as previously described (8, 9). Mean AV plane velocity was computed by dividing AVPD by the duration of ventricular systole. Atrial and ventricular volumes, as well as LV and RV masses, were measured by planimetry. To enable phase-by-phase comparison of KE between individuals with minimal smearing due to different duration of cardiac phases between subjects, KE data were linearly interpolated in time to a reference heartbeat. For the interpolation, the cardiac cycle was divided into the following four phases: ventricular systole, early ventricular diastole (corresponding to the Doppler E wave), diastasis, and late ventricular diastole (atrial systole, Doppler A wave). The start of ventricular systole was defined as the first timeframe in each dataset, triggered to the ECG R wave. The end of ventricular systole was defined by extrapolating the down slope of the aortic flow curve to zero. Figure 1 shows how the end of the E wave was defined from through-plane measurement of transmitral flow. The beginning of the A wave was defined using the same method and marked the end of diastasis. In three of the subjects, the resting heart rate was elevated with subsequent shortening or loss of diastasis, so that the curve extrapolation method was not applicable. In these cases, we visually determined the end of the E wave and the beginning of the A wave. In the two cases without diastasis, only data from the other three cardiac phases were entered in the mean KE calculations. After collating data points in one set with mean values, the duration of each cardiac phase in the reference heartbeat was set equal to the mean phase duration for all subjects. The mean energy was then calculated for each time point. To determine the energy content by volume, the mean energy was divided by the mean volume in each time point. This produced a measure of KE density, which is proportional to the squared blood velocity. The difference in mean energy between the atria was also calculated for each cardiac phase. The increase of KE during early ventricular diastole was measured as the difference between the early diastolic KE and a baseline, as shown in Fig. 2. The baseline was estimated by an exponential decay whose asymptote was zero and whose initial height was chosen to match the KE just before the beginning of the early diastolic peak. The exponential function of the baseline was determined by performing a linear Fig. 1. Definition of time phases from in-plane transmitral flow measurements. The part between 25 and 75% of peak flow of the descending limb of the E wave and the ascending limb of the A wave were extrapolated to zero to define the end of early ventricular diastole and the beginning of late ventricular diastole. Diastasis was defined as the time between the crosses. E, early ventricular diastole; D, diastasis; A, atrial systole, similar to transmitral Doppler echocardiography. J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 The sequence for 4D PC flow measurements has been validated (7). Typical imaging parameters for the four-dimensional (4D) flow acquisitions were: echo time (TE)/repetition time (TR) 3.7/6.3 ms, ␣ 8°, SENSE factor 2, spatial resolution 3 mm isotropic, acquired temporal resolution of 50 ms, and a reconstructed temporal resolution of 25 ms. Typical cine imaging parameters were: TE/TR 2.8/1.4 ms, ␣ 60°, in-plane spatial resolution 1.3 ⫻ 1.3 mm, slice thickness 8.0 mm, no gap, and temporal resolution 30 ms. In total, the scan time for flow measurements and cine imaging was about 50 min. Image analysis and KE calculation. Images were analyzed using an in-house-developed module for the image analysis software Segment (http://segment.heiberg.se) (17). A first-order polynomial fit to stationary tissue was used to compensate for eddy currents and other phase background effects. Velocity aliasing was corrected using phase unwrapping. The velocity in X, Y, and Z directions was squared for each voxel and the three-dimensional velocity vector for each voxel was computed as the square root of the sum of these squared velocities. The resulting velocity magnitude datasets may be broken up into slices to facilitate delineation of the atrial blood pool. Velocity data were therefore sliced with the short-axis cine images as spatial reference. This enabled us to delineate both atria in the cine images, after which the delineations were transferred to the 4D dataset and manually corrected when needed. To calculate the blood mass, the volume of each voxel was multiplied by the blood density, assumed to be 1.05 g/cm3. The KE of the blood within each voxel was then calculated as KE ⫽ ½mv2, where m is the mass of the blood in each voxel, and v is the blood velocity in each voxel. Total atrial KE was then calculated by summation of all voxels within the delineations. The amount of rotational KE in both atria was quantified in each time frame as follows. First, the instantaneous atrial center of mass was found using the atrial delineations. Angular momentum around the center of mass was calculated for each point in time using the velocity and position vectors for each voxel. Total angular momentum was obtained through summation of all voxels. The rotational axis was determined from the center of mass and the direction of angular momentum around this point. The rotational KE (RKE) about this axis was then calculated for each voxel as RKE ⫽ ½mv␣2, where v␣ is the angular velocity component around the computed axis in each voxel. Total rotational KE was then calculated by summation of all voxels in each atrium. The degree of helicity in atrial flow during early ventricular diastole was quantified as follows: the average flow direction was determined from the vector sum of the atrial blood pool. The angle between this average flow direction and the rotational axis was then computed. Values near 0 or 180 degrees indicate a net movement of blood along the rotational axis (helical flow), and values near 90 degrees indicate nonhelical flow. • 1474 Atrial Kinetic Energy in Humans • Arvidsson PM et al. Table 2. Statistical analysis Compared KE peaks Ventricular Systole vs. E Wave Ventricular Systole vs. A Wave E Wave vs. A Wave Left atrium Right atrium ⬍0.001 ⬍0.001 ⬍0.01 ⬍0.001 ⬍0.001 ⬍0.001 P values are shown. Comparison between kinetic energy (KE) peaks in the left and right atrium during ventricular systole, early ventricular diastole (E wave), and late ventricular diastole (A wave), using the Wilcoxon signed-rank test. regression analysis of the middiastolic KE down slope, when the ventricles and atria are stationary. This model thus approximates the exponential decay of KE in the absence of further cardiac movements after ventricular systole. The early diastolic KE increase was then measured as the height of the early diastolic KE peak over the baseline. Visualization of KE. The location of KE in the heart was visualized using the open-source 4D flow visualization software FourFlow (http://fourflow.heiberg.se). The 4D PC-MRI velocity data and anatomical three-chamber, four-chamber, and short-axis images were exported to FourFlow using a plugin for Segment. In the visualization, low KE (⬍0.01 mJ/ml) was gradually faded out, and KE ⬍0.005 mJ/ml was completely masked for clarity, but was included in the total KE calculations. Statistical analysis. Measurements of energies and volumes are expressed as means ⫾ SE, unless stated otherwise. Wilcoxon’s paired nonparametric test was used to test the significance of differences between energy peaks. Linear regression analysis was performed to test the significance of the correlation between AV plane velocity and mean KE, and atrial volumes and KE. Results with a P value ⬍0.05 were considered statistically significant. RESULTS Atrial volumes were within the normal range (20). The mean left atrial (LA) peak volume was 89 ⫾ 5 ml, and 45 ⫾ 2 ml/m2 normalized to body surface area (BSA). The mean right atrial Fig. 3. Mean left (LA) and right (RA) atrial KE over the cardiac cycle for all subjects. Error bars indicate SE. Gray vertical lines differentiate between ventricular systole (S), early ventricular diastole, diastasis, and atrial contraction, respectively. Three energy peaks are seen: 1) ventricular systole, 2) early ventricular diastole, and 3) late ventricular diastole. During ventricular systole, the RA KE is significantly larger than the LA KE. The early diastolic peak was larger in the LA, and the increase in KE from ventricular end systole to early ventricular diastole was larger in the LA. The late diastolic peaks were caused by atrial contraction and did not differ between the atria. J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 Fig. 2. Kinetic energy (KE) of the left and right atrium in one representative subject. Gray boxes indicate actual measurements, and black circles denote hypothetical KE decay if the heart would cease moving at ventricular end systole. To find the exponential function of the decay curve, a linear regression analysis was performed on the middiastolic downward slope, when the ventricles and atria are stationary. The regression curve was then aligned to the downward slope during ventricular end systole. The shaded area between the curves therefore is the amount of KE that is added by cardiac movements during early ventricular diastole. The increase in KE was quantified as the height of the early diastolic peak over the decay curve, indicated by the double-headed arrow. The results of these measurements are shown in Fig. 8. (RA) peak volume was 125 ⫾ 7 ml, and 63 ⫾ 3 ml/m2 normalized to BSA. The ratio between mean RA and mean LA volumes averaged 1.45 ⫾ 0.07. LA KE contained three different energy peaks. A typical subject is shown in Fig. 2. The first peak was found in ventricular systole, the second in early ventricular diastole, and the third in late ventricular diastole. P values for comparisons between peaks are summarized in Table 2. The systolic peak (1.5 ⫾ 0.1 mJ) was smaller than the early diastolic peak (3.5 ⫾ 0.25 mJ, P ⬍ 0.001). Both the systolic and the early diastolic peaks were larger than the late diastolic peak (0.8 ⫾ 0.1 mJ, P ⬍ 0.01 for both). During early ventricular diastole, LA KE increased on average by 3.2 ⫾ 0.27 mJ. Figure 3 shows mean atrial KE for all subjects over the cardiac cycle. To investigate the dependence of atrial KE on atrial volume, the mean KE was divided by the mean atrial volume in each time point. The results, a measure of KE density, are shown in Fig. 4. The mean LA blood KE was 1.1 ⫾ 0.1 mJ and correlated with the mean LA volume, as shown in Fig. 5 (r2 ⫽ 0.38, P ⬍ 0.05). RA KE contained three energy peaks, similarly to the LA (Figs. 2 and 3). However, the systolic peak was larger (3.9 ⫾ 0.3 mJ) than the early diastolic peak (2.9 ⫾ 0.3 mJ, P ⬍ 0.001). Both the systolic and early diastolic peaks were larger than the late diastolic peak (0.9 ⫾ 0.1 mJ, P ⬍ 0.001). During early ventricular diastole, RA KE increased on average by 1.5 ⫾ 0.1 mJ. The mean RA blood KE was 1.7 ⫾ 0.1 mJ and correlated Atrial Kinetic Energy in Humans • 1475 Arvidsson PM et al. Systolic mean KE [mJ] A LA 4 R2 =0.26 y=25x-0.09 P=0.05 3 RA 2 R2 =0.41 y=40x+0.02 P<0.05 1 0 0.00 0.02 0.04 0.06 0.08 AV plane velocity [m/s] Fig. 5. The relationship between atrial volume and KE. Mean LA (solid squares) and RA (open circles) KE by mean LA and RA volume, respectively. The atrial mean KE correlated with the mean atrial volume. These results suggest that mean atrial volume is a strong predictor of atrial blood KE. LA 4 R2 =0.55 y=0.014x-0.3 P<0.01 3 RA 2 R2 =0.49 y=0.018-0.02 P<0.01 1 0 0 50 100 150 200 Systolic peak volume [ml] Fig. 6. The contribution of atrioventricular (AV) plane velocity and peak atrial volume to atrial KE during ventricular systole. To investigate the contribution of AV plane movement to atrial blood KE, AV plane velocity was measured and plotted against the mean KE during ventricular systole. The systolic peak volumes of the atria were also measured and plotted. A: the correlation between AV plane velocity and mean atrial KE during ventricular systole. RA KE correlated with the AV plane velocity, but there was no statistically significant correlation between AV plane velocity and LA KE. B: the correlation between peak atrial volume during ventricular systole and mean atrial KE, with a strong correlation in both atria. Because of the constant-volume properties of the heart, a decrease in ventricular volume must lead to a corresponding increase in atrial volume (2, 5). AV plane movement drives atrial filling (2). It was therefore expected that atrial KE would be determined in part by AV plane velocity. was higher than calculations of squared AV plane velocity times peak atrial blood mass (Fig. 7B). During early ventricular diastole, LA KE increased more than RA KE (P ⬍ 0.001). The increase in LA KE during early ventricular diastole correlated with LV mass (r2 ⫽ 0.28, P ⬍ 0.05; Fig. 8) and with BSA (r2 ⫽ 0.26, P ⬍ 0.05), but not with atrial size. There was no correlation between early diastolic RA KE increase and RV mass, BSA, or atrial size. The location and magnitude of KE in one representative volunteer is shown in Fig. 9. During ventricular systole, LA KE was located mainly at the pulmonary vein inlets, and RA KE was observed in a toroidal pattern with connections to the superior and inferior caval veins. During early ventricular diastole, LA KE was concentrated to the pulmonary vein inlets and surrounding the mitral valve, and the RA KE torus was deformed apically through the tricuspid valve into the RV. During peak filling, the angle between the rotational axis and the average flow direction was 85 ⫾ 16 (mean ⫾ SD) degrees for the LA and 39 ⫾ 23 degrees for the RA (P ⬍ 0.001), indicating helical flow in the RA but not in the LA (Fig. 10). J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 with the mean RA volume, as shown in Fig. 5 (r ⫽ 0.53, P ⬍ 0.01). Comparing the LA and RA energy patterns, the systolic energy peak was larger in the RA (P ⬍ 0.001), and the early diastolic peak was larger in the LA (P ⬍ 0.05). The late diastolic peak did not differ significantly between the atria (P ⫽ 0.17). Mean KE correlated with mean atrial volume over the cardiac cycle for both atria (Fig. 5). During ventricular systole, the downward AV plane velocity correlated with the mean atrial KE (Fig. 6A). Furthermore, the systolic mean KE correlated with the systolic peak volume in both atria (Fig. 6B). The product of downward AV plane velocity and peak atrial volume correlated with systolic mean KE in both atria (P ⬍ 0.01 for LA, P ⬍ 0.001 for RA; Fig. 7A). KE during systole 2 Systolic mean KE [mJ] B Fig. 4. Atrial KE density over the cardiac cycle for all subjects (mean KE/mean volume). Error bars indicate SE. Note that the general shapes of the curves are preserved from Fig. 3, suggesting that volume differences alone do not explain different KE in the atria. LA KE density increases sharply during early ventricular diastole for two reasons. 1) As the left ventricle (LV) relaxes, it generates a pressure gradient that aspirates blood into the ventricle. The increase in LA KE during early ventricular diastole reflects the ability of the LV to generate said pressure gradient. Right ventricular relaxation causes a smaller increase in RA KE, which may be due to the asymmetric configuration and thin wall of the right ventricle (RV). 2) The LV stroke volume exceeds the reservoir volume of the LA. To fill the LV, blood is drawn through the atrium from the pulmonary veins, during which the LA acts as a flow conduit. Because the velocity of a fluid moving through a vessel is inversely proportional to the squared diameter of the vessel, a smaller atrium results in greater blood KE during early ventricular diastole. 1476 Atrial Kinetic Energy in Humans A Arvidsson PM et al. LA 4 Systolic mean KE [mJ] • R2 =0.57 y=0.22x+0.12 P<0.01 3 2 RA R2 =0.64 y=0.24x+0.59 P<0.001 1 0 0 5 10 15 AV plane velocity*peak atrial volume [m/s*ml] 4 3 LA R2 =0.52 y=6.5x+0.39 P<0.01 2 RA R2 =0.56 y=4.9x+1.19 P<0.01 1 0 0.0 0.1 0.2 0.3 0.4 0.5 ½(peak atrial blood mass)*(AV plane velocity)2 [mJ] Fig. 7. A: the contribution of downward AV plane velocity times peak atrial volume to mean atrial KE, during ventricular systole. The combination of these two factors was a better predictor of atrial KE during ventricular systole than either factor by itself (cf. Fig. 6). B: the correlation between squared AV plane velocity times the atrial blood mass and mean atrial KE during ventricular systole. See text for details. Late diastolic KE was concentrated to the regions surrounding the mitral and tricuspid valves, the atrial appendages, and the caval and pulmonary vein inlets. Rotational KE constituted on average 46 ⫾ 7% (SD) of total KE in the LA and 46 ⫾ 8% in the RA (P ⫽ 0.93). The ratio between rotational KE and total KE was not significantly correlated to heart rate (P ⫽ 0.08 for both atria) or to mean total KE (P ⫽ 0.22 for both atria). As shown in Fig. 11, Fig. 8. The correlation between ventricular mass and increase in atrial KE during early ventricular diastole. The increase in LA KE correlated with LV mass. RA KE increase showed no correlation with ventricular mass. Fig. 9. Visualization of the location and magnitude of KE in the heart of one representative volunteer at three time points during the cardiac cycle: at peak ventricular systole (top), at peak early diastolic filling (middle), and during atrial contraction (bottom). Three imaging planes are shown: three-chamber view (3CH, left), four-chamber view (4CH, middle), and a midatrial short-axis view (SAX, right). During ventricular systole, atrial KE is more prominent in the RA than in the LA, where KE is mainly located near the pulmonary vein inlets. LA KE increases sharply during early diastolic filling, which is seen as a streak from the middle of the LA into the center of the LV. Meanwhile, RA KE also increases, but not as much. This increase in KE is likely caused by ventricular recoil. The KE during atrial contraction has a similar location as during early diastolic filling, but the magnitude is smaller. Ao, aorta; Pu, pulmonary trunk. rotational KE increased later than nonrotational KE during systole. The ratio between rotational and nonrotational KE (Fig. 11) increased toward the end of ventricular systole and during diastasis and decreased during early ventricular systole, early ventricular diastole, and atrial contraction. The highest ratios were attained during end-systole and diastasis. DISCUSSION In this article, we present for the first time an analysis of the KE of the LA and RA blood over the whole cardiac cycle in healthy volunteers. RA KE is higher than LA KE, and this is only in part due to higher RA volume compared with the LA. The higher AV plane velocity in the right heart explains part of the remaining difference between the RA and LA. This suggests that ventricular longitudinal contraction is the mechanism responsible for atrial filling and generation of atrial KE and that ventricular diastolic recoil mainly drives LV filling but not RV filling. Relation to earlier studies. Our finding that RA KE is higher than LA KE in all subjects during ventricular systole is consistent with those made by Prec and Katz (31), who used intraventricular pressure measurements together with angiocardiography to approximate the amount of KE involved in LV and RV stroke work and found that KE represents a greater fraction of the total work in the right heart than in the left heart. J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 Systolic mean KE [mJ] B Atrial Kinetic Energy in Humans • 1477 Arvidsson PM et al. Rotational KE Non-rotational KE Ratio RKE/NRKE Left atrium 2.5 0.5 0.0 0.0 0.2 0.4 0.6 0.8 0.0 1.0 Time [fraction of R-R interval] Right atrium 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 Ratio Kim et al. (25) used CMR to estimate the KE of the LV vortex ring during early ventricular diastole and found a mean KE of 0.43 mJ. In their study, the straight flow was not measured, and as a consequence the total LV KE was not quantified, which may explain the lower KE compared with our findings. The diastolic KE of the ventricles was first quantified using 4D PC-MRI in the LV by Eriksson et al. (12) and in the RV by Fredriksson et al. (13). Our measurements of diastolic atrial KE are similar to the findings of both Eriksson et al. and Fredriksson et al. We found peak atrial KE during atrial contraction to be 0.3–2.3 mJ, in good agreement with the 0.25–1.8 mJ described in an earlier echocardiographic study of 1.0 1.0 0.5 KE [mJ] Fig. 10. Helicity of flow during early ventricular diastole. A: the instantaneous angle between net flow and axis of rotation in both atria during peak ventricular filling (mean ⫾ SD). An angle close to 0 or 180 degrees indicates helical flow, and angles near 90 degrees indicate nonhelical flow. The angle was significantly lower in the RA compared with the LA, indicating more helical flow in the RA. B: a streamline visualization of flow across the AV plane in one subject during peak rapid filling of the ventricles. Streamline color indicates KE. Solid white arrows indicate flow direction closest to the viewer, and dashed arrows indicate flow further away. A counterclockwise helix (as viewed from the apex) was seen in RA flow. 1.5 1.5 Ratio KE [mJ] 2.0 2.0 0.2 0.4 0.6 0.8 0.0 1.0 Time [fraction of R-R interval] Fig. 11. Rotational KE (open circles), nonrotational KE (filled squares), and the ratio between the two (gray diamonds). Means ⫾ SE. When total KE decreased, the ratio was found to increase, indicating that rotational flow conserves KE to a greater extent than straight flow does. Systolic peaks in rotational KE occurred after peaks in straight KE, suggesting that straight flow drives rotation. J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 LA KE by Stefanadis et al. (37). Compared with an earlier study where we used 4D PC-MRI to quantify the KE of both ventricles in systole and diastole (6), the present study found atrial KE to be in the same order of magnitude as ventricular KE, although lower during both systole and diastole. At rest, the energy required for LV external stroke work is typically around 1 J (30, 31). At ventricular end systole, the LA blood holds a mean KE of 1.2 mJ, roughly one-thousandth of the energy used for LV stroke work. Mossahebi et al. (28) calculated the pressure work performed by the recoiling chamber during early ventricular diastole and found values in the range of 0.05– 0.2 J, at least one order of magnitude greater than the atrial KE. Therefore, the vast majority of the heart’s energy expenditures are used to generate the pressure necessary to overcome the arterial pressure, and only little is used to create flow. KE, however, is independent of pressure conditions and provides a measure of how much energy that goes into actually moving blood, due to its inertia. Apart from the late diastolic atrial KE, which is caused by atrial contraction, KE in the atria is a consequence of ventricular action. Previous studies have shown that longitudinal pumping provides the major part of LV and RV systolic function (8), that ventricular contraction drives atrial filling (3, 8, 36), and that the LV fills during ventricular diastole by active suction (4, 21). Atrial KE reflects these physiological mechanisms and gives new insights into the relationship between the ventricles and atria. Systolic KE. The increase in atrial KE during ventricular systole may be explained by the mechanism driving atrial 1478 Atrial Kinetic Energy in Humans Arvidsson PM et al. tance at rest. However, given the narrow range of heart rates in this study, it is not possible to rule out an increasing importance during exercise. Diastolic KE. During early ventricular diastole, the LA KE quickly rises to the same approximate level as that of the RA (Fig. 3) and attains a higher KE density (Fig. 4). The increase in LA energy during early ventricular diastole is larger than that of the RA despite the smaller AVPD of the LV. The atria function both as volume reservoirs and flow conduits and contribute to ventricular filling by delivery of three distinct volume fractions: the reservoir volume, the conduit volume, and the atrial stroke volume. The relationship between these volumes has been thoroughly investigated (19, 20, 38). Under normal conditions, the conduit volume flows from the pulmonary and caval veins and through the atria during early ventricular diastole. Such acceleration of atrial blood requires a pressure gradient with lower pressure in the ventricle than in the atrium. The cause of this gradient can be 1) preatrial, 2) atrial, or 3) postatrial/ventricular. Preatrial generation of a pressure gradient presupposes an increase in pulmonary and caval venous blood pressure. Atrial generation of a pressure gradient involves either atrial contraction due to atrial depolarization or release of atrial intramural elastic energy. Postatrial generation of a pressure gradient involves a decrease in ventricular pressure, in other words diastolic suction. A preatrial pressure increase requires buildup of venous pressure. In the supine position, venous return from the legs is increased, and RA filling may therefore be somewhat augmented by caval vein pressure. To the best of our knowledge, there are no physiological explanations as to why pulmonary venous pressure would suddenly increase at the onset of ventricular diastole. Furthermore, such an increase in venous pressure would distend the atrium during early ventricular diastole. The atria, however, decrease in volume during early ventricular diastole. A constant pressure head in the veins would also lead to a progressive increase in total heart volume during diastasis, which is not observed (2, 5). Therefore, a preatrial pressure increase seems an improbable explanation for ventricular filling during early ventricular diastole. An increase in pressure on the atrial level would involve either active or passive contraction. Active atrial contraction is known to occur from the ECG P wave, with atrial repolarization typically occurring simultaneously with the QRS complex. Therefore, the atria are at their electrical resting state during early ventricular diastole, and no active contraction may occur. Passive atrial contraction entails the release of elastic energy, stored in the atrial wall during ventricular systole. Subsequent release of this energy was suggested as the main mechanism for ventricular filling in a previous study (15). However, if atrial elasticity was indeed the mechanism responsible, we would expect to see a correlation between atrial size and KE increase during early ventricular diastole, since a larger atrium can be expected to store more elastic energy. Atrial size could not predict early diastolic KE gain, and RA KE gain was uncorrelated to both RV mass and BSA. Further to this point, atrial elasticity would be expected to cause backflow in the pulmonary and caval veins during early ventricular diastole, which is not observed in healthy individuals. Atrial elasticity is thus incompatible with the atrial conduit volume, which by definition requires influx of venous blood during early ventricular diastole. Indeed, the only time in the cardiac cycle with J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 filling. The total heart volume has been shown to remain nearly constant over the cardiac cycle (2, 5). Within this volume, the movement of the AV plane toward the apex draws blood into the atria (3, 36, 41). Our results show a greater KE increase in the RA than in the LA during ventricular systole (Fig. 3). This finding is consistent with earlier studies that have quantified the contribution of AVPD in left and right ventricular pumping (9). The AVPD is smaller in the left heart than in the right, which partially explains why the LA KE is lower than RA KE during ventricular systole (Fig. 3). Another cause for the lower systolic LA KE compared with the RA is the smaller total volume of the LA. A smaller atrium gives a lower systolic KE, and, as the LA volume is markedly smaller than the RA volume (20), LA KE is lower during ventricular systole. Indexing atrial KE to atrial volume reveals that RA blood has a higher KE density during ventricular systole (Fig. 4). KE density is proportional to the squared blood velocity, and this difference between the atria shows that atrial volume is not the only determinant of atrial KE. As shown in Fig. 6, AV plane velocity and the peak atrial volume are correlated with atrial KE during ventricular systole. The composite of these two factors showed a good correlation with systolic KE, as shown in Fig. 7A. Therefore, we predict that disease affecting ventricular contractility or atrial volumes may influence atrial KE. Figure 7B shows that systolic atrial KE far exceeds the product of squared AV plane velocity and atrial blood mass. Values on the x-axis can be interpreted as the theoretical KE resulting from interaction between the AV plane and the atrial blood pool. Actual measured KE is higher, suggesting that KE is gained not only from cardiac movements and volumes but is also significantly affected by the acceleration of extracardiac blood into the atria. The size and geometric configuration of the pulmonary and caval vein inlets may therefore affect the total atrial KE. All other factors being equal, smaller vessels are more likely to cause high KE in the inflowing blood. RA flow is highly organized during ventricular systole, with inflow from the caval veins contributing to a clockwise (as viewed from the subject’s right side) rotational movement of blood (23). While the net velocity of such rotational flow may be near zero, the total energy content may be high. Indeed, systolic RA KE was found in a toroidal pattern with connections to the superior and inferior caval veins, matching the flow characteristics earlier described. Interestingly, we found the ratio between mean rotational KE and mean total KE to be almost identical between the atria. This may indicate that the anatomically offset configuration of pulmonary and caval veins provides energetically similar conditions in both atria. During periods of energy loss (end-systole and diastasis), rotational KE in both atria decreased slower than nonrotational KE and came to constitute a larger fraction of total KE (the ratio between rotational and nonrotational KE is shown in Fig. 11). This finding shows that rotational flow in the heart conserves energy better compared with nonrotational flow and casts new light upon earlier studies that have discussed atrial flow energetics (14, 23). Increases in systolic rotational KE came with a delay compared with nonrotational KE, suggesting that rotational flow is driven by straight flow during systole. The relationship between total and rotational KE was independent of heart rate and total KE. Considering the total energy expenditure of the ventricles during systole, this finding implies that the “flywheel” mechanism is unlikely to be of major impor- • Atrial Kinetic Energy in Humans Arvidsson PM et al. 1479 circulation and is structurally asymmetric compared with the more cylindrical LV. The myocytes of the RV are also aligned longitudinally to a greater extent than those of the LV, lacking the circumferential “constrictor fibers” necessary for radial stability and generation of high pressures (33). If the RV were to function as a suction pump, the negative pressure gradient would likely cause the RV wall to bulge inward during early diastole. This does not occur in the healthy heart. In contrast to the LV, the RV may instead depend on the greater reservoir volume of the RA for filling during early ventricular diastole. During ventricular systole, this greater volume is subject to a larger acceleration due to the longer AVPD of the right heart. At the beginning of ventricular diastole, the RA therefore contains a larger volume of blood with a higher KE than the LA. Under such conditions, it is possible that opening of the tricuspid valve is sufficient to achieve RV filling. The KE of the inflowing blood could thus contribute to the distension of the RV and the basal movement of the AV plane, rather than being a consequence of RV recoil. Meanwhile, the open tricuspid valve allows the RV to slide past the atrial blood as the AV plane moves basally, thereby encompassing the atrial blood and moving it into the ventricle essentially without imparting velocity onto it. Because this means that the blood remains relatively stationary in relation to the body and the CMR scanner, it does not lead to an increase in KE. In short, the marked increase in LA KE is compatible with the model of active LV suction during early ventricular diastole. Thus, LA KE may be understood as a consequence of, and not a cause for, LV filling. Conversely, RA KE does not show evidence for ventricular suction. It seems reasonable that RV filling is optimized for the low-pressure conditions of the pulmonary circulation, where a lack of RV wall thickness and structural rigidity is compensated for by a higher systolic RA KE. This RA KE spills over into ventricular diastole and may be used to fill the relaxing RV. Physiological implications. With increasing exercise comes a shortening of ventricular diastole, which in turn requires more efficient ventricular filling. KE constitutes a greater share of the total work in the right heart during rest and may become increasingly important during work (6). In light of this relative abundance of RA KE and lack of RV diastolic suction, RA KE may provide a mechanistic explanation to right ventricular filling, especially so at a higher cardiac output. The higher end-systolic RA KE compared with the LA means that there is already a greater movement of blood in the RA that does not need to be accelerated in early ventricular diastole. Instead, when the tricuspid valve opens because of declining RV pressure, RA blood can fill the ventricle by merit of its relatively high KE. In comparison, the LA blood is more stationary at the end of ventricular systole. It is likely that both LA and RA KE will be conserved to a greater extent during exercise, since there is no diastasis during which KE can dissipate. During early ventricular diastole, flow in the RA was more helical than in the LA (Fig. 10), which may be a consequence of redirection of rotational flow from systole. This may be a possible mechanism to facilitate RV filling through conservation of atrial KE. Helical flow is rotational flow with a net direction along the rotational axis, and this finding appears consistent with energy conservational redirection of flow (23). Conserving LA rotation through conversion into helical flow J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 slight backflow in the pulmonary veins is during active atrial contraction. Thus, while atrial elasticity may contribute somewhat to ventricular filling, it is probably not a major factor in LV filling. The third possible explanation for a pressure gradient involves a decrease in ventricular pressure, or ventricular suction. Such a mechanism is compatible with the declining atrial volume during early ventricular diastole and with inflow from the pulmonary and caval veins. Ventricular suction gives a mechanistic explanation to the conduit volume. It is also compatible with a true diastasis, with both the ventricles and the atria at a “resting state” with no change in volume after the rapid filling phase. The rapid increase in LA KE during early ventricular diastole therefore appears more consistent with ventricular suction than with an increase in venous or atrial pressure. Earlier studies have also provided evidence for diastolic suction in the LV (1, 4, 21, 28). The cellular mechanism behind diastolic suction lies in elastic proteins that generate a restoring force after myocyte shortening below slack length (18, 32), possibly contributing to the untwisting of the ventricle during early ventricular diastole (1). Ventricular myocytes are macroscopically organized into bands with significant differences in direction between the LV and RV (10, 29, 33). This organizational structure may be important for the ability of the ventricles to generate suction and may also be important for differences between LV and RV suction. We found that RA KE does not increase as much as LA KE during early ventricular diastole. It has been shown that the myocardium contains elastic proteins that give rise to ventricular elastic recoil during early diastole (1, 16, 18, 32). Because these proteins are evenly distributed and integral parts of the cardiac muscle of both ventricles, ventricular mass may be considered a measure of the total amount of proteins with capacity to generate elastic recoil. We therefore hypothesized a correlation between ventricular mass and the increase in atrial KE during early ventricular diastole, which would imply that diastolic atrial KE is generated by ventricular recoil. As shown in Fig. 8, there was a correlation between LV mass and LA KE increase. This suggests that suction arising from the LV accounts for about one-quarter of the variability in LA KE increase in early ventricular diastole. Factors other than LV suction may account for the other three-fourths of the variability in LA KE increase. A correlation between LV mass and LA KE increase could also be explained if both factors were dependent on an external factor, such as body size. If this was the explanation, we would expect to find a similar correlation between RV mass and RA KE increase, since the healthy heart maintains a balance between LV and RV (35). There was no correlation between RV mass and RA KE increase, implying that RV filling depends on other mechanisms. The lack of correlation between RV wall mass and KE generation during early diastole leads us to believe that structural stability, and not simply the amount of myocardium available, affects the amount of ventricular recoil. We suggest that proper structural scaffolding (sufficiently thick and symmetrical myocardial wall) is important for elastic recoil to occur. In a previous study where myocytes were axially compressed in the absence of sideways support, the thick filaments tended to bulge sideways, hampering the ability of the myocyte to generate a restoring force (18). The RV has a thin wall because of the lower working pressures of the pulmonary • 1480 Atrial Kinetic Energy in Humans Arvidsson PM et al. filling but may instead depend more on the combination of atrial reservoir function and atrial KE for sufficient filling. Rotational blood flow contained about one-half of the total atrial KE and was found to conserve KE to a greater extent than nonrotational flow did. Rotational flow in the right atrium was redirected into helical flow during early ventricular diastole, which may be a mechanism that facilitates RV filling through conservation of atrial KE. ACKNOWLEDGMENTS We thank Christoffer Green for help with FourFlow, Ann-Helen Arvidsson for help with data collection, and Katarina Steding-Ehrenborg for help with data analysis. GRANTS This study was supported by grants from the Swedish Research Council, Sweden (2008-2461, 2008-2949, 2011-3916), National Visualization Program and Knowledge Foundation, Sweden (2009-0080), the Swedish Heart and Lung Foundation, Sweden, the Medical Faculty at Lund University, Sweden, and the Region of Scania, Sweden. DISCLOSURES No conflicts of interest are declared by the authors. AUTHOR CONTRIBUTIONS Author contributions: P.M.A., E.H., M.C., and H.A. conception and design of research; P.M.A., J.T., and M.C. performed experiments; P.M.A., J.T., and M.C. analyzed data; P.M.A., M.C., and H.A. interpreted results of experiments; P.M.A. prepared figures; P.M.A. drafted manuscript; P.M.A., J.T., E.H., M.C., and H.A. edited and revised manuscript; P.M.A., J.T., E.H., M.C., and H.A. approved final version of manuscript. REFERENCES 1. Bell SP, Nyland L, Tischler MD, McNabb M, Granzier H, LeWinter MM. Alterations in the determinants of diastolic suction during pacing tachycardia. Circ Res 87: 235–240, 2000. 2. Bowman AW, Kovacs SJ. Assessment and consequences of the constantvolume attribute of the four-chambered heart. Am J Physiol Heart Circ Physiol 285: H2027–H2033, 2003. 3. Bowman AW, Kovacs SJ. Left atrial conduit volume is generated by deviation from the constant-volume state of the left heart: a combined MRI-echocardiographic study. Am J Physiol Heart Circ Physiol 286: H2416 –H2424, 2004. 4. Brecher G. Critical review of recent work on ventricular diastolic suction. Circ Res 6: 554 –566, 1958. 5. Carlsson M, Cain P, Holmqvist C, Stahlberg F, Lundback S, Arheden H. Total heart volume variation throughout the cardiac cycle in humans. Am J Physiol Heart Circ Physiol 287: H243–H250, 2004. 6. Carlsson M, Heiberg E, Toger J, Arheden H. Quantification of left and right ventricular kinetic energy using four dimensional intracardiac magnetic resonance imaging flow measurements. Am J Physiol Heart Circ Physiol 302: H893–H900, 2012. 7. Carlsson M, Toger J, Kanski M, Bloch KM, Stahlberg F, Heiberg E, Arheden H. Quantification and visualization of cardiovascular 4D velocity mapping accelerated with parallel imaging or k-t BLAST: head to head comparison and validation at 1.5 T and 3 T (Abstract). J Cardiovasc Magn Reson 13: 55, 2011. 8. Carlsson M, Ugander M, Heiberg E, Arheden H. The quantitative relationship between longitudinal and radial function in left, right, and total heart pumping in humans. Am J Physiol Heart Circ Physiol 293: H636 –H644, 2007. 9. Carlsson M, Ugander M, Mosen H, Buhre T, Arheden H. 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 292: H1452–H1459, 2007. 10. Coghlan C, Hoffman J. Leonardo da Vinci’s flights of the mind must continue: cardiac architecture and the fundamental relation of form and function revisited. Eur J Cardio-Thoracic Surg 29: S4 –S17, 2006. 11. Engblom H, Steding K, Carlsson M, Mosén H, Hedén B, Buhre T, Ekmehag B, Arheden H. Peak oxygen uptake in relation to total heart J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 may be less important for cardiac function, since the LV work capacity (and hence suction generation) greatly exceeds the KE involved in LA flow. Because KE is directly related to velocity, the results shown in Fig. 3 also mean that the movement of blood in the atria during the cardiac cycle never reaches zero. This is important for two reasons. First, moving blood is less prone to thrombus formation, and, second, the conservation of KE may be important for cardiac pumping, especially during exercise (22). Thus, the present study adds data to the current debate regarding the importance of KE in the looped heart (24, 39, 40). Furthermore, the KE of the healthy heart can be expected to increase with greater cardiac dimensions, contributing to higher exercise capacity (35), in contrast with the failing heart in which the exercise capacity is not related to the total heart volume (11). It is important to note that, while KE is related to momentum, it does not have a direction. Linear momentum is calculated as the product of mass and velocity, which produces a vector. Therefore, whether conservation of the linear momentum of blood may be beneficial for cardiac pumping depends on the direction of momentum. If a portion of blood is moving with high velocity toward a vessel wall, it holds high linear momentum but is potentially harmful for the endothelium and may cause energy loss as a consequence of turbulent flow. Conversely, KE is a nondirectional measure of the amount of work invested in the acceleration of blood, and as such does not consider the direction of flow. Limitations. 4D PC data collection relies on averaging hundreds of heartbeats. Beat-to-beat variations in flow are cancelled out, and only macroscopic flow features are preserved. Turbulence is one of the reasons for KE decay, and, due to the above, the KE of microscopic turbulent flow features was not directly measured. The turbulent flows left out of the calculations are by definition not recurring flow features and therefore interact randomly with blood flow patterns. The impact of these interactions was not quantified. Because of the limited temporal resolution of the 4D data, the isovolumic contraction and relaxation periods are not always clearly distinguishable. This may result in some smearing of the mean values during the beginning and end of ventricular systole but does not affect measures of KE peaks. Apart from AV plane velocity, ventricular wall mass, and atrial volume, several factors may further influence atrial KE. These factors are ventricular elasticity and stiffness (26, 27), preload and central venous pressure, anatomical differences in the caval and pulmonary vein inlets, and size and configuration of the mitral and tricuspid annuli. Left ventricular twisting and untwisting appears to play some role in LV elastic recoil, although the degree of contribution is unclear (1). Subjects may also vary in physiology due to circadian rhythms, hydration, previous food intake, and menstrual cycles. These factors were not investigated further and may be important considerations for future studies. In conclusion, this study has quantified the biatrial kinetic energies over the cardiac cycle and showed that RA KE predominates during ventricular systole, whereas LA KE increases sharply at the onset of ventricular diastole. This constitutes new evidence that the LV is more dependent on the mechanism of diastolic suction for filling compared with the RV. The RV does not appear to rely upon diastolic suction for • Atrial Kinetic Energy in Humans 12. 13. 14. 15. 16. 17. 18. 20. 21. 22. 23. 24. 25. 26. Arvidsson PM et al. 1481 27. Kovacs SJ. Solution of the “inverse problem of diastole” via kinematic modeling allows determination of ventricular properties and provides mechanistic insights into diastolic heart failure. Conf Proc IEEE Eng Med Biol Soc 2009: 2354 –2357, 2009. 28. Mossahebi S, Shmuylovich L, Kovacs SJ. The thermodynamics of diastole: kinematic modeling-based derivation of the P-V loop to transmitral flow energy relation with in vivo validation. Am J Physiol Heart Circ Physiol 300: H514 –H521, 2011. 29. Notomi Y, Popovic ZB, Yamada H, Wallick DW, Martin MG, Oryszak SJ, Shiota T, Greenberg NL, Thomas JD. Ventricular untwisting: a temporal link between left ventricular relaxation and suction. Am J Physiol Heart Circ Physiol 294: H505–H513, 2008. 30. Opie L. Contractile performance of the intact heart. In: Heart Disease A Textbook of Cardiovascular Medicine, edited by Braunwald E, Zipes DP, and Libby P. Philadelphia, PA: Saunders, 2001, p. 462–468. 31. Prec O, Katz LN. Determination of kinetic energy of the heart in man. Am J Physiol 159: 483–491, 1949. 32. Preetha N, Yiming W, Helmes M, Norio F, Siegfried L, Granzier H. Restoring force development by titin/connectin and assessment of Ig domain unfolding. J Muscle Res Cell Motil 26: 307–317, 2006. 33. Sheehan F, Redington A. The right ventricle: anatomy, physiology and clinical imaging. Heart 94: 1510 –1515, 2008. 34. Spencer MP, Greiss FC. Dynamics of ventricular ejection. Circ Res 10: 274 –279, 1962. 35. Steding K, Engblom H, Buhre T, Carlsson M, Mosén H, Wohlfart B, Arheden H. Relation between cardiac dimensions and peak oxygen uptake (Abstract). J Cardiovasc Magn Reson 12: 8, 2010. 36. Steding-Ehrenborg K, Carlsson M, Stephensen S, Arheden H. Atrial aspiration from pulmonary and caval veins is caused by ventricular contraction and secures 70% of the total stroke volme independent of resting heart rate and heart size. Clin Physiol Funct Imaging In press. 37. Stefanadis C, Dernellis J, Lambrou S, Toutouzas P. Left atrial energy in normal subjects, in patients with symptomatic mitral stenosis, and in patients with advanced heart failure. Am J Cardiol 82: 1220 –1223, 1998. 38. Tseng WYI, Liao TY, Wang JL. Normal systolic and diastolic functions of the left ventricle and left atrium by cine magnetic resonance imaging. J Cardiovasc Magn Reson 4: 443–457, 2003. 39. Watanabe H, Sugiura S, Hisada T. The looped heart does not save energy by maintaining the momentum of blood flowing in the ventricle. Am J Physiol Heart Circ Physiol 294: H2191–H2196, 2008. 40. Watanabe H, Sugiura S, Hisada T. Reply to “Letter to the editor: ‘Postulated functional advantages of a looped as opposed to a linearly arranged heart’.” Am J Physiol Heart Circ Physiol 298: H727–H727, 2010. 41. Waters EA, Bowman AW, Kovács SJ. 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 288: H848 –H853, 2005. J Appl Physiol • doi:10.1152/japplphysiol.00932.2012 • www.jappl.org Downloaded from http://jap.physiology.org/ by 10.220.33.2 on May 3, 2017 19. volume discriminates heart failure patients from healthy volunteers and athletes (Abstract). J Cardiovasc Mag Res 12: 74, 2010. Eriksson J, Dyverfeldt P, Engvall J, Bolger AF, Ebbers T, Carlhall CJ. Quantification of presystolic blood flow organization and energetics in the human left ventricle. Am J Physiol Heart Circ Physiol 300: H2135– H2141, 2011. Fredriksson AG, Zajac J, Eriksson J, Dyverfeldt P, Bolger AF, Ebbers T, Carlhall CJ. 4-D blood flow in the human right ventricle. Am J Physiol Heart Circ Physiol 301: H2344 –H2350, 2011. Fyrenius A, Wigstrom L, Ebbers T. Three dimensional flow in the human left atrium. Heart 86: 448 –455, 2001. Grant C, Bunnell IL, Greene DG. The reservoir function of the left atrium during ventricular systole. An angiocardiographic study of atrial stroke volume and work. Am J Med 37: 36 –43, 1964. Granzier HL, Irving TC. Passive tension in cardiac muscle: contribution of collagen, titin, microtubules, and intermediate filaments. Biophys J 68: 1027–1044, 1995. Heiberg E, Sjogren J, Ugander M, Carlsson M, Engblom H, Arheden H. Design and validation of segment: freely available software for cardiovascular image analysis (Abstract). BMC Med Imaging 10: 1, 2010. Helmes M, Trombitás K, Granzier HL. Titin develops restoring force in rat cardiac myocytes. Circ Res 79: 619 –626, 1996. Jarvinen VM, Kupari MM, Hekali PE, Poutanen VP. Assessment of left atrial volumes and phasic function using cine magnetic resonance imaging in normal subjects. Am J Cardiol 73: 1135–1138, 1994. Jarvinen VM, Kupari MM, Hekali PE, Poutanen VP. Right atrial MR imaging studies of cadaveric atrial casts and comparison with right and left atrial volumes and function in healthy subjects. Card Radiol 191: 137– 142, 1994. Katz LN. The role played by the ventricular relaxation process in filling the ventricle. Am J Physiol 95: 542–553, 1930. Kilner PJ, Henein MY, Gibson DG. Our tortuous heart in dynamic mode–an echocardiographic study of mitral flow and movement in exercising subjects. Heart Vessels 12: 103–110, 1997. Kilner PJ, Yang GZ, Wilkes AJ, Mohiaddin RH, Firmin DN, Yacoub MH. Asymmetric redirection of flow through the heart. Nature 404: 759 –761, 2000. Kilner PJ. Letter to the editor: “Postulated functional advantages of a looped as opposed to a linearly arranged heart.” Am J Physiol Heart Circ Physiol 298: H726 –H726, 2010. Kim WY, Walker PG, Pedersen EM, Poulsen JK, Oyre S, Houlind K, Yoganathan A. Left ventricular blood flow patterns in normal subjects: a quantitative analysis by three-dimensional magnetic resonance velocity mapping. J Am Coll Cardiol 26: 224 –238, 1995. Kovacs SJ, Mcqueen DM, Peskin CS. Modelling cardiac fluid dynamics and diastolic function. Philos Trans Royal Soc A Math Phys Eng Sci 359: 1299 –1314, 2001. •