| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Physics and Medical Technology and 2Department of Cardiology, University Medical Center, VU 1007 MB Amsterdam, The Netherlands
Submitted 5 November 2003 ; accepted in final form 9 January 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
left ventricular mechanical asynchrony; healthy human heart; post-systolic shortening; relaxation
A characteristic timing parameter of cardiac contraction is the time to onset of (circumferential) shortening, which can be measured noninvasively with MRI tagging (39). Another characteristic timing parameter is the time to peak (circumferential) shortening. For patients with a left bundle branch block, a marked delay between the septum and the lateral wall has been reported. Both onset of shortening and peak shortening are early in the septum and late in the lateral wall, indicating a discoordinate and hence inefficient contraction (9, 27). Several other studies have shown that ischemic regions continue shortening after aortic valve closure (postsystolic shortening) (10, 31, 34, 35), which leads to the occurrence of peak shortening after aortic valve closure.
Mapping these cardiac contraction times in the normal human heart may give valuable insight into normal contraction patterns and may serve as a reference for interpreting data of patients considered for CRT and patients with ischemia. Time of peak shortening has been reported by Fonseca et al. (14) with a temporal resolution of 35–45 ms. However, onset time of shortening maps have been reported only in dogs (39).
In this study we used a new method for fast acquisition of MRI myocardial tagging images (41). This allows myocardial tagging and strain analysis with a high temporal resolution of 14 ms. Normal, healthy volunteers were studied to determine the spatial distributions of both the onset time of shortening (Tonset) and the time of peak shortening (Tpeak). The relation between Tonset and Tpeak was studied, with the hypothesis that in the normal heart a later onset of shortening leads to a later time of peak shortening. In addition, the relation between Tpeak and the time of aortic valve closure was studied to determine the presence of postsystolic shortening in healthy subjects.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
Seventeen healthy subjects (mean age 42 ± 11 yr; 11 men, 6 women) with no history of cardiac disease, normal ECG, ejection fraction >50%, and blood pressures <160/90 mmHg were studied after informed consent was obtained according to our institutional guidelines.
Image Acquisition
Imaging was performed on a 1.5-T whole body scanner (Magnetom Sonata, Siemens, Erlangen, Germany) using a phased-array receiver coil. Myocardial tagging images with high temporal resolution (14 ms) were obtained with steady-state free precession (SSFP) imaging using the linearly increasing start-up angle (LISA) approach (28, 41). Complementary tagging (complementary spatial modulation of magnetization, CSPAMM) was used for improved strain analysis (13, 21). Five short-axis image planes were acquired from base to apex, four image series per plane: horizontally and vertically tagged images, with both sinusoidal and inverted sinusoidal tagging for CSPAMM. To minimize breath-hold misregistration among the four series, a multiple breath hold scheme was applied, in which data were acquired during brief expiration breath holds of four heartbeats each, which were interleaved by 6-s pauses to inhale and exhale (11, 22). The total acquisition time for a single image plane was 
4.5 min. The tag spacing was 7 mm, and the readout direction was orthogonal to the tag lines. The other imaging parameters can be found in Table 1. Examples of the complementary tagged images are shown in Fig. 1 for both horizontally and vertically tagged images at three different times in the cardiac cycle.
|
|
To determine the timings of the mitral and aortic valves, a long-axis cine in the three-chamber view was acquired. For this purpose, SSFP imaging with viewsharing was used to obtain a temporal resolution of 14 ms in a single breath hold. Conventional short-axis cine images were acquired to assess the ejection fraction of the left ventricle (LV). Imaging parameters are given in Table 1. The acquisition duration of all tagging cines and anatomic cines was 45–55 min/subject.
Postprocessing
Myocardial strains. CSPAMM images were computed from the complementary tagged image series. In <10% of the cases, a small shift was applied in the tag direction (absolute value of 1.5 ± 0.5 pixels) to correct for misregistration. Harmonic magnitude images (29), in which the image intensity is proportional to the tag contrast, were calculated from the CSPAMM images and used to draw endo- and epicardial contours. The myocardial deformation was calculated as described previously (21). Because the LISA approach leads to an inappropriate signal-to-noise ratio in the first image, the deformation was calculated from the second image (at 39 ms after the ECG R wave) onward.
Circumferential strain 
c was calculated from the Lagrangian strain tensor as the percent change in length of a small line segment in the circumferential direction (15)
 | (1) |
c curves represent the shortening of the underlying myocardial fibers, because the fibers at the midwall are oriented mainly in the circumferential direction (39). The first data point of the c curve was discarded, because the LISA approach leads to a low signal-to-noise ratio in the tagged image acquired directly after the application of the tagging pattern.
Mapping timing of contraction. Tonset and Tpeak relative to the ECG R wave were determined for each of the six segments in each slice. Tonset was defined as the beginning of the down slope of the c curve and was determined by an automated fitting algorithm as described in the APPENDIX. Values for Tonset were regarded as missing for segments in which the goodness of fit, R, was <0.9.
In analogy to the QRS width, the onset of shortening width (OS width) and peak shortening width (PS width) within a subject were defined. The OS width and PS width are given by the time needed for 20–90% of the LV segments to start shortening (38) or reach peak shortening, respectively.
From the Tonset map, the two-dimensional onset of shortening delay vector (OS delay vector) was determined, as a measure for the spatiotemporal distribution of the intraventricular asynchrony, as described by Wyman et al. (Ref. 38, where this vector is called the activation delay vector). Similarly, a peak shortening delay vector (PS delay vector) was derived from the Tpeak map. The magnitude of such a vector indicates the time difference between one side of the LV and the other, and its direction points from the earliest segment to the latest segment. The direction of the vector is reported with respect to a reference vector pointing from the septum to the lateral wall. Positive angles express a turn in the counterclockwise direction. For calculation of the OS delay vector, missing values for Tonset were replaced by using bilinear interpolation from the neighboring segments.
Timing of valves. From the long-axis cine in the three-chamber view, the times of the aortic valve opening (Tavo), aortic valve closure (Tavc), and mitral valve opening (Tmvo) were assessed. Opening time was defined as the time relative to the ECG R wave of the first cine frame on which the valves were recognized as open. Closure time was defined as the time of the first frame on which the small, dark backflow jet associated with valve closure was observed. The mitral valves were normally already closed on the first cine frame; hence the moment of closure of these valves could not be determined. The duration of isovolumic relaxation (IVR) was calculated as Tmvo – Tavc. Figure 2 summarizes the timing parameters described above.
|
Relation between timings. The relation between Tonset and Tpeak was tested by calculating the correlation between Tonset and Tpeak of the 30 segments, which were averaged across the subjects. Furthermore, it was tested whether there were segments in which peak shortening was reached significantly later than aortic valve closure, defining postsystolic shortening. The duration of postsystolic shortening was quantified as the duration between aortic valve closure and time of peak shortening (Tpeak – Tavc) averaged over all segments with significant postsystolic shortening and over all subjects. Similarly, the amount of postsystolic shortening was c(Tpeak) – c(Tavc) averaged over the same segments, where c(Tpeak) – c(Tavc) was taken as zero when peak shortening occurred before aortic valve closure for an individual subject. Finally, the asynchrony in Tpeak (quantified as either the PS width or the magnitude of the PS delay vector) was correlated with the duration of IVR. A larger asynchrony in Tpeak is likely to prolong the duration of IVR, thereby decreasing the efficiency of the cardiac cycle. Because both the asynchrony in Tpeak and the duration of IVR are likely to be associated with heart rate, the partial correlation coefficient was calculated, adjusting for heart rate (3).
Statistics. Data are presented as means ± SD. The nonparametric Wilcoxon test was used for paired tests, and the Spearman correlation coefficient was used to study the relations between timings. Differences were regarded as statistically significant at P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
|
Onset Time of Shortening
The onset time of shortening could be determined in >98% of the cases, because only 10 of 17 x 30 segments were rejected because of a moderate goodness of fit. On average, the onset of shortening started in the AL segment at midventricular level at 53 ± 15 ms after the ECG R wave, then spread out over the lateral segments and the apex, and finally reached the basis of the IS and IN segments at 83 ± 17 ms (see Fig. 4). The average OS width was 32 ± 8 ms (range 18–41 ms), indicating a fast activation. Although regional differences were small, there was a distinct spatial distribution of the onset time of shortening. The apex started 9 ± 8 ms earlier than the basis (P < 0.05). The AN, AL, and IL segments (averaged over slices) were significantly earlier than the IN, IS, and AS segments, respectively (see Table 2). This was reflected by the mean OS delay vector pointing from the lateral wall to the septum at an angle of –147 ± 42° with an amplitude of 15 ± 13 ms (range 2–39 ms), revealing a consistent pattern in the normal heart of early shortening in the lateral wall. In one subject the OS delay vector pointed from the IS to the AL segment, opposite to the normal direction, but only with an amplitude of 2 ms, indicating a spatially homogeneous onset of shortening. In another subject, the OS delay vector pointed from IN to AN, whereas in all other subjects the OS delay vector pointed from one of the lateral segments to one of the septal segments.
|
|
Time of Peak Shortening
Peak shortening occurred earliest in the septal and anterior segments at midventricular levels and latest in the IL segment, as shown in Fig. 4. The average PS width for all subjects was 121 ± 22 ms (range 84–168 ms), which is more than three times larger than the OS width. The spatial distribution of the time to peak shortening was largely opposite to the spatial distribution of the onset of shortening, the septal segments being earlier than the lateral segments (Table 2). This was reflected by the average PS delay vector, which pointed from the AS to the IL segment. The angle of the average PS delay vector was –17 ± 22° (range –43 to 52°), and the magnitude was 57 ± 25 ms (range 26–95 ms).
Relation Between Tonset and Tpeak
We found a negative correlation between Tonset and Tpeak (r =–0.59, P = 0.001), indicating that regions with an earlier onset of shortening have a longer duration of shortening (see Fig. 5).
|
Postsystolic Shortening
Postsystolic shortening was observed in 13 of the 30 segments, located predominantly in the IL and AL segments (see Fig. 4B). The average duration of postsystolic shortening was 58 ± 14 ms, which is as much as 16 ± 4% of systolic duration. During this postsystolic shortening, the amount of shortening increased by 3.1 ± 0.9% from 16.9 ± 1.2% at Tavc to 20.0 ± 1.5% at Tpeak, a relative increase of 19 ± 6%.
Relation Between Asynchrony in Tpeak and Duration of IVR
Adjusting for heart rate, the partial correlation coefficient between the duration of IVR and the PS width was r = 0.62 (P < 0.02). This indicates that the duration of IVR is related to the amount of asynchrony in the time to peak shortening (see Fig. 6). When the magnitude of the PS delay vector was used as a measure for the asynchrony in peak shortening, no relation was found with the duration of IVR (see DISCUSSION).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
In this study, timing of the cardiac contraction was mapped by means of high-temporal-resolution MRI tissue tagging and strain analysis. The time needed for the LV to start shortening (OS width, 32 ± 8 ms) corresponds well to the time needed for electrical activation of the LV endocardium (30 ms) (12, 16). Apparently, the transmural delay, which makes the total electrical activation duration of the LV 60 ms, is not seen in the maps of the onset of shortening at midwall.
The electrical activation sequence of the heart proceeds from the septum to the lateral wall and therefore does not correspond to the mechanical sequence of the onset of shortening, which starts in the lateral wall and ends in the inferior and IL wall. It has been suggested, based on computer simulations, that the delay between electrical and mechanical activation (E-M delay) is spatially heterogeneous to ensure a mechanically synchronous contraction (19). Variations in E-M delay could arise from either differences in load or differences in initial sarcomere length. A higher load causes an increase in E-M delay, because a higher force must be developed by the myocytes before shortening will occur. A larger sarcomere length leads to a decrease in E-M delay, because of a mechanism that is known as length-dependent activation and is related to the increased calcium sensitivity for larger sarcomere lengths (20, 30, 36). Because the load is assumed to be homogeneously distributed over the fibers (5), we hypothesize that the earlier onset of shortening in the lateral wall may be caused by a local increase in initial fiber length due to the atrial contribution to filling. The lateral wall may be stretched more than the septum by the atrial filling, because it is thinner and because the septum is subject to counterbalancing effects from the left and right ventricular blood pools.
The high temporal resolution of 14 ms enabled us to determine small differences in Tonset of 7 ms with significance (Table 2). With a typical temporal resolution of 30–40 ms used in many tagging studies, the detection of these small differences is far more difficult and may at best be seen only in large populations.
Time to Peak Shortening
Time of peak shortening was earlier in the septum than in the lateral wall and hence showed an opposite spatial pattern compared with the onset time of shortening. As a result, duration of shortening was longest in the lateral wall, for which it is known that peak strains are largest (25). Thus the extra amount of shortening in the lateral wall is at least partly due to a longer shortening duration.
The amount of asynchrony in the peak shortening (PS width, 121 ± 22 ms) was about three times larger than the amount of asynchrony in the onset of shortening but was of the same order of magnitude as the duration of IVR (102 ± 15 ms). Fonseca et al. (14) used as an asynchrony measure for Tpeak the absolute difference between Tpeak for a segment and the mean over all segments and averaged this difference over all segments. When this measure was used, our subjects had a regional asynchrony in Tpeak of 44.6 ± 7.4 ms, which is in line with the values Fonseca et al. reported: 37.5 ± 7 ms for a younger group (19–26 yr) and 46.8 ± 11.5 ms for an older group (60–74 yr). The temporal resolution used by Fonseca et al. was 35 ms. Abraham et al. (1) used Doppler strain rate imaging to measure asynchrony in the onset of relaxation, which was defined as the zero crossing of the strain rate at the end of contraction. They reported a low variability in the onset of regional relaxation (maximum regional difference was only 41 ms), which is much less than the asynchrony we found in Tpeak. The different findings may be explained by the fact that Abraham et al. studied the rate of the longitudinal strain whereas we examined circumferential shortening and by the fact that peak shortening is sometimes followed by a period of relatively constant strain before relaxation occurs (see Fig. 3).
Postsystolic Shortening
Several segments in the lateral wall and in the basal regions demonstrated significant postsystolic shortening. Recently, the occurrence of postsystolic shortening in healthy subjects was reported by Voigt et al. (37), using Doppler longitudinal strain rate imaging. For regions with postsystolic shortening, they reported time of peak postsystolic shortening of 61 ms (median) after Tavc, which agrees with the mean duration of postsystolic circumferential shortening found in the present study (58 ± 14 ms). However, Voigt et al. reported postsystolic shortening predominantly in the apical and basal segments of the septal, AS, and anterior walls, but not in the lateral wall.
The amount of postsystolic shortening is quite substantial, because its duration was about one-sixth of the systolic duration with a relative increase in strain of 19 ± 6% after aortic valve closure. The physiological function of this postsystolic shortening remains unclear. However, the presence of postsystolic shortening in healthy hearts complicates the interpretation of postsystolic shortening as a sign of ischemia.
Relation between Tonset and Tpeak
The negative correlation between Tonset and Tpeak was not expected. It seems that regions with an earlier onset of shortening have a later peak shortening and therefore a longer duration of shortening. The longer shortening duration for the regions with early onset of shortening is another indication for our hypothesis that prestretch due to the atrial contribution is heterogeneous and the largest in the lateral wall.
Relation Between Asynchrony in Tpeak and Duration of IVR
The results suggested that the duration of IVR is related to the asynchrony in Tpeak, as may be expected. It is interesting to note that this relation was found only when the asynchrony was quantified as the PS width (time needed for 20–90% of the LV to reach peak strain). When the magnitude of the PS delay vector was used, no correlation was found with the duration of IVR. It should be noted that the PS delay vector is a two-dimensional vector that takes into account the asynchrony of opposite segments only. Apparently, the findings imply for the peak shortening that asynchrony in the longitudinal direction, as well as asynchrony other than between opposite segments, play an important role in the duration of IVR.
Limitations
This study used two-dimensional strain calculations to derive the circumferential shortening. Therefore, the effect of through-plane motion during the acquisition was not taken into account. Because the strain variations in the longitudinal direction are relatively small (25), the effect of through-plane motion is limited.
The LISA approach used to combine the fast SSFP imaging with myocardial tagging leads to a low signal-to-noise ratio in the first cine image after the ECG R wave. Therefore, the first image that could be used for strain analysis had a delay of 39 ms with respect to the R wave. The deformation occurring before this time, in the first part of systole, cannot be studied with our methods. Nevertheless, this early deformation appears to be small, because the first strain measured is typically less than only 1% (see Fig. 3). The earliest detectable Tonset depends also on the delay of the first cine image. Onsets of shortening earlier than 39 ms are clipped to 39 ms by the Tonset estimation algorithm, because the down slope of the strain curve starts at 39 ms. This may lead to an underestimation for the mechanical asynchrony. Although the exact onset of shortening cannot be determined in early regions, the fact that they are early (Tonset 
39 ms after the ECG R wave) will be detected as long as shortening continues during the first few cine images.
The closure of the mitral valve was normally too early to be detected on the prospectively ECG-triggered MRI cines. Therefore, it was not possible to measure the duration of the isovolumic contraction phase and to study the relation between this duration and the asynchrony in the onset of shortening.
This study focused on healthy subjects. The value of the mechanical timing parameters used for characterization of disease and for predicting the response to treatment must be established in the future.
In conclusion, onset of shortening spread quickly over the LV, with a consistent spatial pattern. Onset of shortening started in the AL segment halfway between apex and base and reached the IN and IS segments at the basis last. Tpeak was less homogeneous compared with Tonset and showed spatially a reversed pattern, the septum being earlier than the lateral wall. Postsystolic shortening, defined as the occurrence of the peak shortening after aortic valve closure, was observed in 13 of the 30 cardiac segments, mainly in the lateral and basal segments. Knowledge of the spatial pattern of contraction timing in the normal heart may serve as a reference in detecting abnormalities in myocardial contraction due to intraventricular conduction defects or ischemia.
|
APPENDIX: ALGORITHM FOR ESTIMATION OF ONSET TIME OF SHORTENING |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
Input
The algorithm uses the circumferential strain curve c(t) and the harmonic magnitude curve as input.
Fit Model
The onset of shortening is determined by fitting the following model
 | (2) |
The model consists of two line segments, of which the first line segment for t T0 represents the period before onset of shortening and the second segment represents the period of shortening. Slope a1 must be positive, corresponding to the zero-shortening or prestretch stage, and a2 must be negative, corresponding to shortening. The constant a0 accounts for a small strain offset that is caused by variations in the tag distance due to field inhomogeneity-related gradients during the application of the tagging pattern. tend is the end point of the data that are used in the fit.
Algorithm
The algorithm consists of the following steps. 1) Exclude akinetic segments, which are defined as segments with minimum c >–4%. Continue with step 2 only if the segment is not akinetic.
2) Automatically select data points to which L is fit. a) The first data point is removed, because of its low signal-to-noise ratio in the LISA approach (41), and the point (0,0) is added. b) The end point tend of the time interval is the moment after peak strain rate at which the strain rate declines under 70% of peak strain rate. The strain rate is calculated from a filtered copy of the c curve (filtering of c is done with a moving average filter of 5 sample widths, to reduce noise sensitivity). c) The data range is forced to have at least four points from t = 0 (only relevant in patients, with very brief durations of shortening in early activated segments).
3) Fit L to data of the selected time interval. a) The harmonic magnitude is used as a weight factor to account for the dependence of the accuracy in c on the tag contrast [the weight factor for (0,0) is the same as that for the first data point]. b) L is fit to the selected c data range, for T0 values t1 T0 < tend, where t1 is the time delay of the first strain data point (39 ms in our case) and the step size is 1 ms. c) The onset time of shortening Tonset is the value of T0 that minimizes the residual error (sum of squares).
Implementation
The algorithm was implemented in Matlab (version 6.5; MathWorks, Natick, MA) with the lsqlin function for the constrained linear least-squares fit mentioned in step 3b.
|
ACKNOWLEDGMENTS |
|---|
GRANTS
This work was supported by the Netherlands Heart Foundation, grant number 2000B220.
|
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 |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIX: ALGORITHM FOR...REFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Nasiraei-Moghaddam and M. Gharib Evidence for the existence of a functional helical myocardial band Am J Physiol Heart Circ Physiol, January 1, 2009; 296(1): H127 - H131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Delgado, C. Ypenburg, R. J. van Bommel, L. F. Tops, S. A. Mollema, N. A. Marsan, G. B. Bleeker, M. J. Schalij, and J. J. Bax Assessment of Left Ventricular Dyssynchrony by Speckle Tracking Strain Imaging: Comparison Between Longitudinal, Circumferential, and Radial Strain in Cardiac Resynchronization Therapy J. Am. Coll. Cardiol., May 20, 2008; 51(20): 1944 - 1952. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Marcus, C. T.-J. Gan, J. J.M. Zwanenburg, A. Boonstra, C. P. Allaart, M. J.W. Gotte, and A. Vonk-Noordegraaf Interventricular Mechanical Asynchrony in Pulmonary Arterial Hypertension Left-to-Right Delay in Peak Shortening Is Related to Right Ventricular Overload and Left Ventricular Underfilling. J. Am. Coll. Cardiol., February 19, 2008; 51(7): 750 - 757. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Marcassa, R. Campini, E. Verna, L. Ceriani, and P. Giannuzzi Assessment of cardiac asynchrony by radionuclide phase analysis: Correlation with ventricular function in patients with narrow or prolonged QRS interval Eur J Heart Fail, May 1, 2007; 9(5): 484 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Lucats, B. Ghaleh, X. Monnet, P. Colin, A. Bize, and A. Berdeaux Conversion of post-systolic wall thickening into ejectional thickening by selective heart rate reduction during myocardial stunning Eur. Heart J., April 1, 2007; 28(7): 872 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ashikaga, B. A. Coppola, B. Hopenfeld, E. S. Leifer, E. R. McVeigh, and J. H. Omens Transmural Dispersion of Myofiber Mechanics: Implications for Electrical Heterogeneity In Vivo J. Am. Coll. Cardiol., February 27, 2007; 49(8): 909 - 916. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Sanderson Systolic and Diastolic Ventricular Dyssynchrony in Systolic and Diastolic Heart Failure J. Am. Coll. Cardiol., January 2, 2007; 49(1): 106 - 108. [Full Text] [PDF]
|
|
|
|
|
|
M. J.W. Gotte, T. Germans, I. K. Russel, J. J.M. Zwanenburg, J. T. Marcus, A. C. van Rossum, and D. J. van Veldhuisen Myocardial Strain and Torsion Quantified by Cardiovascular Magnetic Resonance Tissue Tagging: Studies in Normal and Impaired Left Ventricular Function J. Am. Coll. Cardiol., November 21, 2006; 48(10): 2002 - 2011. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. P. Sengupta, J. Korinek, M. Belohlavek, J. Narula, M. A. Vannan, A. Jahangir, and B. K. Khandheria Left Ventricular Structure and Function: Basic Science for Cardiac Imaging J. Am. Coll. Cardiol., November 21, 2006; 48(10): 1988 - 2001. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Sanderson Assessment of ventricular dyssynchrony: global or regional function? Eur. Heart J., October 2, 2006; 27(20): 2380 - 2381. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Knaapen, M. J. W. Gotte, W. J. Paulus, J. J. M. Zwanenburg, P. A. Dijkmans, R. Boellaard, J. T. Marcus, J. W. R. Twisk, C. A. Visser, A. C. van Rossum, et al. Does Myocardial Fibrosis Hinder Contractile Function and Perfusion in Idiopathic Dilated Cardiomyopathy? PET and MR Imaging Study. Radiology, August 1, 2006; 240(2): 380 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J.M. Zwanenburg, M. J.W. Gotte, J. T. Marcus, J. P.A. Kuijer, P. Knaapen, R. M. Heethaar, and A. C. van Rossum Propagation of Onset and PeakTime of Myocardial Shortening in Time of Myocardial Shortening in Ischemic Versus Nonischemic Cardiomyopathy: Assessment by Magnetic Resonance Imaging Myocardial Tagging J. Am. Coll. Cardiol., December 20, 2005; 46(12): 2215 - 2222. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vonk-Noordegraaf, J. T. Marcus, C. T. Gan, A. Boonstra, and P. E. Postmus Interventricular Mechanical Asynchrony Due To Right Ventricular Pressure Overload in Pulmonary Hypertension Plays an Important Role in Impaired Left Ventricular Filling Chest, December 1, 2005; 128(6_suppl): 628S - 630S. [Full Text] [PDF]
|
|
|
|
|
|
A. Vonk-Noordegraaf, J. T. Marcus, C. T. Gan, A. Boonstra, and P. E. Postmus Interventricular Mechanical Asynchrony Due To Right Ventricular Pressure Overload in Pulmonary Hypertension Plays an Important Role in Impaired Left Ventricular Filling Chest, December 1, 2005; 128(6_suppl): 628S - 630S. [Full Text] [PDF]
|
|
|
|
|
|
H. Ashikaga, S. R. Mickelsen, D. B. Ennis, I. Rodriguez, P. Kellman, H. Wen, and E. R. McVeigh Electromechanical analysis of infarct border zone in chronic myocardial infarction Am J Physiol Heart Circ Physiol, September 1, 2005; 289(3): H1099 - H1105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. M. Zwanenburg, M. J. W. Gotte, J. P. A. Kuijer, M. B. M. Hofman, P. Knaapen, R. M. Heethaar, A. C. van Rossum, and J. T. Marcus Regional timing of myocardial shortening is related to prestretch from atrial contraction: assessment by high temporal resolution MRI tagging in humans Am J Physiol Heart Circ Physiol, February 1, 2005; 288(2): H787 - H794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Jérôm. Garot The study of diastole by tagged MRI: are we nearly there yet? Eur. Heart J., August 2, 2004; 25(16): 1376 - 1377. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, estimated Tonset; 
, Tpeak. The vertical lines denote, from left to right, the moment of aortic valve opening (dashed), aortic valve closure (solid), and mitral valve opening (dashed). IS, infero-septal; AS, antero-septal; AN, anterior; AL, antero-lateral; IL, infero-lateral; IN, inferior.
P < 0.05. C: Tonset averaged over all subjects, shown as a function of slice position and cardiac segment. The error bars indicate SD and are shown for the earliest and the latest segment only. D: Tpeak averaged over all subjects and shown for all slices and myocardial segments. The solid and dashed lines indicate Tavc ± SD.
