Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print September 11, 2007, 10.1148/radiol.2451061791
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Material
Right arrow All Versions of this Article:
2451061791v1
245/2/567    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Husmann, L.
Right arrow Articles by Alkadhi, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Husmann, L.
Right arrow Articles by Alkadhi, H.
(Radiology 2007;245:567-576.)
© RSNA, 2007

Technical Developments

Coronary Artery Motion and Cardiac Phases: Dependency on Heart Rate—Implications for CT Image Reconstruction1

Lars Husmann, MD, Sebastian Leschka, MD, Lotus Desbiolles, MD, Tiziano Schepis, MD, Oliver Gaemperli, MD, Burkhardt Seifert, PhD, Philippe Cattin, PhD, Thomas Frauenfelder, MD, Thomas G. Flohr, PhD, Borut Marincek, MD, Philipp A. Kaufmann, MD, and Hatem Alkadhi, MD

1 From the Institute of Diagnostic Radiology (L.H., S.L., L.D., T.F., B.M., H.A.) and Cardiovascular Center (T.S., O.G., P.A.K.), University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich, Switzerland; Department of Biostatistics (B.S.) and Center for Integrative Human Physiology (P.A.K.), University of Zurich, Zurich, Switzerland; Computer Vision Laboratory, Swiss Federal Institute of Technology, ETH Zurich, Zurich, Switzerland (P.C.); and Computed Tomography CTE PA, Siemens Medical Solutions, Forchheim, Germany (T.G.F.). From the 2006 RSNA Annual Meeting. Received October 17, 2006; revision requested December 19; revision received January 26, 2007; final version accepted March 7. Supported by the National Center of Competence in Research, Computer Aided and Image Guided Medical Interventions of the Swiss National Science Foundation and by the Georg und Bertha Schwyzer-Winiker Stiftung, Switzerland. Address correspondence to H.A. (e-mail: hatem.alkadhi{at}usz.ch).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
This study had institutional review board approval; written informed consent was obtained. The purpose was to prospectively determine the heart rate (HR) dependency of three-dimensional (3D) coronary artery motion by incorporating into analysis the durations of systole and diastole. Thirty patients (seven women, 23 men; mean age, 56.6 years ± 12.7 [standard deviation]; HR: 45–100 beats per minute) underwent electrocardiographically gated 64-section computed tomographic (CT) coronary angiography to determine coronary motion velocities at bifurcation points. Significance of velocity differences (P < .05) was determined by using analysis of variance for repeated measures and Bonferroni post hoc tests. HR dependency was determined by using linear regression analysis. HR significantly affected 3D coronary motion (r = 0.47, P < .009) through nonproportional shortening of systole and diastole (r = –0.82, P < .001), leading to percentage reconstruction interval shifts of coronary velocity troughs and peaks (P < .01). Results suggest that image reconstruction algorithms at CT coronary angiography be adapted to the individual patient's HR.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2451061791/DC1

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Coronary arteries undergo heterogeneous movement and deformations throughout the cardiac cycle (1) that cause motion artifacts on computed tomographic (CT) images when the velocity exceeds the temporal resolution of the scanner (2). Early studies on coronary artery motion used the varying locations of coronary artery bifurcation points on biplane coronary cine angiograms as a basis for motion analysis (3–5). However, these studies suffered from inaccurate measurements because of depth-dependent amplification in projection imaging and the mathematic assumptions that were required to calculate three-dimensional (3D) coordinates from two-dimensional representations. More recently, studies involving electron-beam CT (6–8) and magnetic resonance imaging (5,9,10) have investigated coronary artery motion by following labeled positions on coronary arteries throughout the cardiac cycle, but these studies restricted their two-dimensional analysis to the transverse plane, and the effects of through-plane movements were not mentioned in their reports. Vembar et al (11) studied 3D coronary artery motion with four–detector row CT; however, the results of this study were limited by the fact that patients with high heart rates were not included.

Motion artifacts in electrocardiographically gated cardiac CT are usually minimized by reconstructing images at phases of near quiescence in mid-diastole called diastases (6–8,11–15). With increasing heart rates, however, nonproportional shortening of systole and diastole occurs, and the phase of diastasis progressively diminishes and eventually disappears (16,17). A relationship between heart rate and image quality at multi–detector row CT coronary angiography has been noted by several authors (14,18,19), but neither the effect of different heart rates on coronary artery motion nor the effect of nonproportional shortening of systole and diastole with regard to reconstruction intervals has been investigated so far. Our hypothesis was that motion artifacts should be minimized when motion of the coronary arteries is minimal. Thus, the purpose of our study was to prospectively determine the heart rate dependency of 3D coronary artery motion by incorporating into the analysis the durations of systole and diastole.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Authors who are not employees of or consultants for Siemens Medical Solutions (Forchheim, Germany) had control of inclusion of any data and information that might present a conflict of interest for the author (T.G.F.) who is an employee of that company.

Patients
Between May and December of 2005, 115 consecutive patients (43 women, 72 men; mean age, 59.9 years ± 14.2; age range, 17–82 years) underwent CT coronary angiography. The patients were suspected of having coronary artery disease (n = 89) or had a history of myocardial infarction with recurrent angina (n = 26). Exclusion criteria were as follows: allergy to iodinated contrast agents, renal insufficiency (creatinine level > 120 µmol/L), pregnancy, nonsinus rhythm, and hemodynamic instability. Patients with significant coronary artery stenoses (stenoses ≥ 50% of lumen diameter) (n = 15), patients with vessel wall calcification without significant stenoses (stenoses < 50% of lumen diameter) (n = 54), patients with abnormal origin and course of the coronary arteries (n = 5), and patients with heart rate variability of more than 10 beats per minute during scanning (n = 11) were excluded from the study to minimize interfering factors affecting coronary artery motion. Thus, the final study group included 30 patients (seven women, 23 men; mean age, 56.6 years ± 12.7; age range, 17–76 years). The study protocol was approved by the local ethics committee, and written informed consent was obtained from all patients.

CT Data Acquisition and Postprocessing
Seventeen patients (57%) were receiving oral ß-receptor blockers as part of their baseline medications. No additional ß-blockers were administered prior to CT. All examinations were performed by using a 64-section CT scanner (Sensation 64; Siemens Medical Solutions) and the following parameters: detector collimation, 32 x 0.6 mm; section collimation, 64 x 0.6 mm (z-flying focal spot); gantry rotation time, 330 msec; pitch, 0.2; tube potential, 120 kV; and tube current–time product, 650 mAs. A bolus of 80 mL of iodixanol (Visipaque 320 [320 mg iodine per milliliter]; GE Healthcare, Buckinghamshire, England) followed by 30 mL of saline solution was continuously injected into an antecubital vein through a 20-gauge catheter at a flow rate of 5 mL/sec. Bolus tracking was performed with a region of interest in the ascending aorta, and image acquisition was automatically started 5 seconds after the signal attenuation reached a predefined threshold of 140 HU. Electrocardiography-synchronized CT data sets were retrospectively reconstructed throughout the cardiac cycle in 10% steps of the R-R interval with a section thickness of 1 mm and an increment of 0.8 mm (medium soft-tissue convolution kernel, B30f). The "adaptive cardio volume approach," which uses a single-segment reconstruction algorithm at heart rates lower than 65 beats per minute (temporal resolution, 165 msec) and a two-segment reconstruction algorithm at heart rates of 65 beats per minute or higher (temporal resolution, 83 msec), was employed (20).

Determination of Length of Systole and Diastole
The beginnings of systole and diastole were determined in each patient so that we could relate the physiologic changes of the phases of the cardiac cycle caused by different heart rates (21) to coronary artery motion. The time interval in which systole and diastole began was individually defined in each patient on the basis of aortic valve opening and closure (21,22), respectively. To do this, we reconstructed double-oblique parallel planes through the aortic annulus in 10% steps of the R-R interval, as previously shown (23). Opening and closing of the aortic valve were noted by two radiologists (L.H., with 3 years of experience in cardiovascular imaging, and H.A., with 7 years of experience) in consensus.

Coronary Artery Motion in 3D
Eleven landmarks, named according to the guidelines of the American Heart Association (24) (Fig 1), at coronary bifurcation points were identified in each patient in every data set in 10% steps of the R-R interval by one radiologist (L.H.). To determine the accuracy of the measurements, interobserver variation (between L.H. and H.A.) was calculated for the first three patients for all measured coordinates. Owing to the low interobserver variation (0.45 mm ± 0.91 [standard deviation]), measurements in the remaining 27 patients were performed by only one reader (L.H.). Measurements were manually made by positioning a cursor at the exact center of each landmark, and the spatial x-, y-, and z-coordinates were recorded by using dedicated software (Watsyn; Research Systems, Boulder, Colo). From these coordinates, the 3D route of coronary artery motion was geometrically calculated in each phase (p) of the R-R interval by using the following equation:

Formula
where pp(n) represents a given phase of the R-R interval and pp(n–1) represents the previous phase.


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Definition of 11 landmarks at coronary bifurcation points according to international guidelines (24): segment 1, ostium of right coronary artery (RCA); segment 1/2, origin of right ventricular branch; segment 2/3, origin of acute marginal branch; segment 3/4, origin of posterior descending branch; segment 5, ostium of left main artery; segment 5/11, origin of left circumflex (LCX) artery; segment 6/9, origin of first diagonal branch; segment 7/10, origin of second diagonal branch; segment 11/12, origin of obtuse marginal branch; segment 13/14, origin of first posterolateral branch; and segment 13, origin of second posterolateral branch.

 
The velocity of coronary artery motion of each landmark was calculated as the quotient of the route and the length of the time interval, which was computed as 60 divided by the mean heart rate divided by 10. With this approach, for example, the time interval 20%–30% refers to the velocity of coronary bifurcation points between 20% and 30% of the R-R interval. We used a method similar to that used in a previous study (6) in that all time instances were expressed as percentages of the R-R interval to enable comparisons of velocities across the 30 patients, with the cardiac cycle defined as beginning at the onset of the R wave. Thus, 10% relative time intervals at a mean heart rate of 45–100 beats per minute corresponded to absolute time intervals of 133–60 msec. For illustration purposes, 3D coronary artery motion was plotted as a function of the percentage time intervals.

On the basis of these calculations, the following variables were obtained: (a) The velocity of coronary arteries, defined as the velocity of the RCA (mean velocity in segments 1, 1/2, 2/3, and 3/4), the left anterior descending (LAD) artery (mean velocity in segments 5, 5/11, 6/9, and 7/10), and the LCX artery (mean velocity in segments 5/11, 11/12, 13/14, and 13) throughout the R-R interval; (b) the velocity of landmarks, defined as the velocity of each bifurcation point throughout the entire R-R interval; (c) velocity within time intervals, defined as the velocity of all bifurcation points averaged for all patients in each 10% interval, enabling the detection of velocity peaks and troughs; (d) heart rate dependency of coronary artery motion, determined as the relation of velocity peaks and troughs in each patient to the individual heart rate; (e) heart rate dependency of the mid-diastolic velocity trough (as an indicator for the optimal window for image reconstruction), determined as the relation of the lowest mid-diastolic coronary artery motion velocity to the individual heart rate and determined as the relation of the width of the mid-diastolic velocity trough in each patient to the individual heart rate (the latter defined as the difference of x-axis values in percentages of the R-R interval, where velocities were lower than a cutoff value of 20 mm/sec); and (f) heart rate dependency of minimal systolic and diastolic coronary motion (as an indicator for the heart rate above which image reconstructions should be performed in systole), determined as the relation of the intraindividually calculated differences of the lowest velocity in systole and diastole to the individual heart rate.

Statistical Analysis
Quantitative variables were expressed as means ± standard deviations. Interobserver agreement for the determination of the length of systole and diastole was calculated with Cohen {kappa} statistics (25) and interpreted according to the guidelines of Landis and Koch (26). The relationship between the length of systole and heart rate was analyzed with Spearman rank-order correlation coefficients.

The velocities of coronary artery motion did not display normal distribution (Q-Q plots in Figure E1, http://radiology.rsnajnls.org/cgi/content/full/2451061791/DC1). Therefore, velocity measurement data were log transformed before analysis to meet assumptions of normality and were expressed as geometric means ± geometric standard deviations, as previously described (27). With this method, the data set approached normal distribution (Figure E1, http://radiology.rsnajnls.org/cgi/content/full/2451061791/DC1), and powerful statistical analyses could be performed. Because the log-transformed data were approximately normally distributed, the geometric mean of velocities was comparable to the median, and the geometric standard deviation can be considered a multiplicative measure of variation around the median.

Linear regression with 95% individual prediction intervals was used to assess the relationship between heart rate and the length of systole and diastole, the relationship between heart rate and the shifts of velocity peaks and troughs, the relationship between heart rate and the widths of the mid-diastolic velocity troughs, and the relationship between heart rate and the log ratio of the smallest velocities in diastole and systole (the latter on a log scale). All tests were two sided, and P < .05 was considered to indicate a statistically significant difference.

Differences between 3D coronary artery motion velocities, velocities of landmarks, and velocities within time intervals were assessed by using analysis of variance for repeated measures. Significance was calculated for all three analyses (P < .001); therefore, Bonferroni post-hoc tests could be used for multiple pairwise comparisons. Corresponding P values were Bonferroni corrected and thus are considered statistically significant for P < .05. All statistical analyses were conducted by using software (SPSS, version 12.0.1; SPSS, Chicago, Ill) (B.S.).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
Mean heart rate during scanning was 66.1 beats per minute ± 17.1 (range, 45–100 beats per minute), with a heart rate variability of 3.3 beats per minute ± 2.3 (range, 0.7–7.4 beats per minute). Two hundred ninety-six landmarks were available for analysis in all 10 time intervals. Fourteen landmarks were absent because of anatomic variations of distal branches, and 20 landmarks (RCA [n = 12], LAD artery [n = 1], and LCX artery [n = 7]) could not be adequately localized in systole because of severe motion artifacts; the data sets were therefore discarded.

Determination of Length of Systole and Diastole
Interobserver agreement was good ({kappa} = 0.79). Immediate agreement between both observers was achieved in 58 cases (97%); in the remaining two cases (3%), consensus reading was required to discriminate the beginning of diastole between 30% and 40% and 40% and 50% of the R-R interval. The beginning of systole (ie, the opening of the aortic valve) occurred in all patients at 0% of the R-R interval. The beginning of diastole (ie, the closing of the aortic valve) took place at variable intervals, from 20% to 50% of the R-R interval. Although the length of the entire cardiac cycle shortened from 1.33 seconds at a heart rate of 45 beats per minute to 0.60 second at a heart rate of 100 beats per minute and the length of systole shortened from 0.4 to 0.36 second, the proportional length of systole as a percentage of the R-R interval significantly increased with the higher heart rate (r = 0.85, P < .001; Fig 2), with the transition taking place at 20% with the low heart rate (45 beats per minute) and at 50% with the high heart rate (100 beats per minute) (mean transition point for all patients, 34.8% ± 7.8). Thus, with increasing heart rate, the percentage duration of systole lengthened from 30% to 60% of the cardiac cycle.


Figure 2
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Graph shows that Spearman rank correlation coefficients for associations between length of systole as percentage of R-R interval and heart rate (in beats per minute [bpm]) in 30 patients were statistically significant (r = 0.85, P < .001).

 
Three-dimensional Coronary Artery Motion
Velocity of coronary arteries.—Mean velocity of all segments in all phases and patients was 26.8 mm/sec ± 2.6 (range, 1.0–333.6 mm/sec). The fastest velocity was measured in the RCA (mean, 35.8 mm/sec ± 2.8; range, 1.0–333.6 mm/sec; P < .001). Second-highest velocities were found in the LCX artery (mean, 24.9 mm/sec ± 2.5; range, 1.0–249.0 mm/sec); these velocities were significantly faster (P < .001) than velocities in the LAD artery (mean, 20.2 mm/sec ± 2.3; range, 1.0–131.9 mm/sec). Movies 1 and 2 (http://radiology.rsnajnls.org/cgi/content/full/2451061791/DC1), which were generated from 64-section CT coronary angiography data, illustrate coronary artery motion in two perpendicular projections.

Velocity of landmarks.—The fastest velocity across all heart rates and all time intervals was found at the origin of the acute marginal branch (segment 2/3: 48.0 mm/sec ± 3.0), and the second-highest velocity was found at the origin of the right ventricular branch of the RCA (segment 1/2: 41.9 mm/sec ± 2.8, no significant difference) (Table). The velocities of these landmarks were significantly higher than that of any other segment (P < .001). The lowest velocity was found at the origin of the LCX artery (segment 5/11: 19.9 mm/sec ± 2.3); this velocity was significantly lower than the velocities of all segments in the RCA and the LCX artery (P < .005) but was not significantly different from the velocities of other LAD artery landmarks.


View this table:
[in this window]
[in a new window]

  		Geometric Mean Velocities of 3D Coronary Artery Motion in 30 Patients

 
Velocity within time intervals.—A velocity trough (13.0 mm/sec ± 2.3) was found between 60% and 70% of the R-R interval (corresponding to the slow ventricular filling phase); velocities in this interval were significantly lower than those in all other time intervals (P < .001). A second velocity trough (26.1 mm/sec ± 2.3) was found between 20% and 30% of the R-R interval (corresponding to the reduced ventricular ejection phase); velocities in this interval were significantly lower than those in neighboring time intervals (10%–20%: 34.5 mm/sec ± 2.3, P < .001 and 30%–40%: 31.7 mm/sec ± 2.2, P < .05).

Peak coronary artery motion velocities (42.5 mm/sec ± 2.3) were found between 0% and 10% of the R-R interval (corresponding to the early rapid ventricular ejection phase); velocities in this interval were higher than those in neighboring time intervals (10%–20%: 34.5 mm/sec ± 2.3, P < .02 and 90%–100%: 32.4 mm/sec ± 2.6, P = not significant). A second velocity peak (31.7 mm/sec ± 2.2) was detected between 30% and 40% of the R-R interval (corresponding to the early rapid ventricular filling phase); velocities in this interval were again higher than those in neighboring time intervals (20%–30%: 26.1 mm/sec ± 2.3, P < .04 and 40%–50%: 28.5 mm/sec ± 2.6, P = not significant). A third velocity peak (corresponding to the phase of atrial contraction) could be visually appreciated between 80% and 90% for segments 1/2, 2/3, and 13; however, this peak disappeared when velocity data were averaged across all segments (Table). Coronary artery motion velocities at all bifurcation points in all patients are shown in Figure 3. For comparison with values from previous studies (6–9,11), arithmetic velocities of coronary artery motion are provided in Table E1 (http://radiology.rsnajnls.org/cgi/content/full/2451061791/DC1).


Figure 3
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Graph shows mean coronary artery motion velocities of each landmark (colored lines) and of all landmarks (black line) in 30 patients plotted against time in percentages of R-R interval. Peak coronary artery motion velocities with statistically significant differences were found between 0% and 10% and between 30% and 40% of the R-R interval. An additional velocity peak can be visually appreciated for segments 1/2, 2/3, and 13 between 80% and 90% of the R-R interval. Two velocity troughs with statistically significant differences are present between 20% and 30% and 60% and 70% of the R-R interval. Mean percent reconstruction phase (34.8%) indicating end of systole and beginning of diastole is indicated by black vertical line. Seg. = segment.

 
Heart rate dependency of coronary artery motion.—With increasing heart rate, the mean velocity of all coronary arteries (r = 0.47, P < .009) and of each coronary artery significantly increased (RCA: r = 0.40, P < .04; LAD artery: r = 0.59, P < .002; LCX artery: r = 0.41, P < .03). Velocity calculations in relation to heart rate are illustrated for all coronary arteries in Figure 4 and for each coronary artery in Figure 5.


Figure 4
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Graph shows coronary artery motion velocities of all landmarks according to heart rate. For illustration purposes, patients were grouped according to heart rate: lower than 60 beats per minute (bpm) (n = 10), between 60 and 75 beats per minute (n = 10), and higher than 75 beats per minute (n = 10). Velocities were plotted against percentages of the R-R interval. Arrows indicate shifts of first velocity trough and second velocity peak and decrease of width of the second velocity trough with increasing heart rate. Time interval of first velocity peak showed no changes with different heart rates. Range of shift throughout all patients and heart rates was larger than the average shift for the 10 patients constituting each patient subgroup: The first velocity trough shifted with increasing heart rate from the interval of 10%–20% to 50%–60% of the cardiac cycle (r = 0.51, P < .01), the second velocity peak shifted from 20%–30% to 60%–70% (r = 0.77, P < .001), and the second velocity trough shifted from 40%–50% to 70%–80% (r = 0.63, P < .001).

 

Figure 5A
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: Coronary artery motion velocities of (a) RCA, (b) LAD artery, and (c) LCX artery according to heart rate. For illustration purposes, patients were grouped according to heart rate: lower than 60 beats per minute (bpm) (n = 10), between 60 and 75 beats per minute (n = 10), and higher than 75 beats per minute (n = 10). Shifts to time intervals later in the cardiac cycle for velocity peaks and troughs and decrease in width of mid-diastolic velocity trough with increasing heart rate can be appreciated for each coronary artery.

 

Figure 5B
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: Coronary artery motion velocities of (a) RCA, (b) LAD artery, and (c) LCX artery according to heart rate. For illustration purposes, patients were grouped according to heart rate: lower than 60 beats per minute (bpm) (n = 10), between 60 and 75 beats per minute (n = 10), and higher than 75 beats per minute (n = 10). Shifts to time intervals later in the cardiac cycle for velocity peaks and troughs and decrease in width of mid-diastolic velocity trough with increasing heart rate can be appreciated for each coronary artery.

 

Figure 5C
View larger version (11K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5c: Coronary artery motion velocities of (a) RCA, (b) LAD artery, and (c) LCX artery according to heart rate. For illustration purposes, patients were grouped according to heart rate: lower than 60 beats per minute (bpm) (n = 10), between 60 and 75 beats per minute (n = 10), and higher than 75 beats per minute (n = 10). Shifts to time intervals later in the cardiac cycle for velocity peaks and troughs and decrease in width of mid-diastolic velocity trough with increasing heart rate can be appreciated for each coronary artery.

 
Increases in heart rate were accompanied by significant time shifts of the two velocity troughs. The early velocity trough shifted with increasing heart rate from the interval of 10%–20% to 50%–60% of the cardiac cycle (r = 0.51, P < .01). The second velocity trough shifted from 40%–50% to 70%–80% when heart rate increased (r = 0.63, P < .001). Incorporating the analysis of the start and end of systole and diastole (described above) revealed the first velocity trough to be located in systole and the second to be located in diastole in each patient.

The early velocity peak between 0% and 10% of the R-R interval showed no significant temporal shift with increasing heart rate (mean, 3.0% ± 6.0; r = 0.0; P = not significant). The second velocity peak shifted from 20%–30% to 60%–70% when heart rate increased (r = 0.77, P < .001). Incorporating the analysis of the start and end of systole and diastole (described above) revealed the first velocity peak to be located in systole and the second velocity peak to occur in diastole in each patient.

Heart rate dependency of mid-diastolic velocity trough.—With increasing heart rate, the lowest coronary artery velocity in the mid-diastolic trough significantly increased (range, 4.4–31.5 mm/sec; r = 0.75; P < .001). Furthermore, with increasing heart rate, the width of the mid-diastolic trough significantly decreased (r = –0.82, P < .001) and the trough eventually disappeared (ie, velocity in diastole exceeded 20 mm/sec).

Heart rate dependency of minimal systolic and diastolic coronary motion.—With increasing heart rate, the lowest velocities in diastole continuously increased while the minimum velocities in systole remained almost unchanged at a relatively high velocity level (Figs 4, 5). The intraindividually calculated differences of the lowest velocity in systole and diastole in each patient showed a significant negative correlation with increasing heart rate (r = –0.82, P < .001) (Fig 6). The cutoff heart rate at which velocity differences between systole and diastole turned to negative values (indicating lower velocities in systole than in diastole) was 83 beats per minute for all coronary arteries. The cutoff heart rate for each coronary artery was 84 beats per minute for the RCA, 78 beats per minute for the LAD artery, and 80 beats per minute for the LCX artery.


Figure 6
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Linear regression plot with 95% individual prediction interval (dotted lines) of intraindividually calculated ratio of systolic and diastolic minimal velocity against heart rate in all patients and all coronary arteries indicates a significant negative correlation (y = –0.013x + 1.12, r = –0.82, P < .001). A value greater than 1 on y-axis indicates lowest coronary velocities to be in diastole. A value less than 1 on y-axis indicates lowest coronary velocities to be in systole. At heart rates greater than 83 beats per minute (bpm) (see interception of horizontal line with regression line), the lowest coronary artery motion velocity for all segments shifted from diastole to systole (lower 95% individual prediction interval at 58 beats per minute [see interception of horizontal line with lower individual prediction interval line]; upper 95% individual prediction interval could not be calculated [see no interception of horizontal line with upper individual prediction interval line within limits of the scale]).

 
The effect of increasing heart rate on image quality at 64-section CT coronary angiography and the duration of systole and diastole regarding reconstruction intervals are illustrated in three patients in Figure 7.


Figure 7
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7: Transverse images from 64-section CT coronary angiography data set in three patients with heart rates of 45 beats per minute (bpm) (top row), 63 beats per minute (middle row), and 85 beats per minute (bottom row) demonstrate influence of increasing heart rate on image quality, reconstruction intervals, and duration of systole and diastole. With increasing heart rate, best image quality (large white squares) shifts to later percentage phases of the R-R interval (indicated by number in lower left corner of each image). This is accompanied by an increase in the length of systole and a decrease in the length of diastole. Boxes in lower right corners show opening and closing of aortic valve throughout R-R interval; small white squares show percentage phase of aortic valve closure.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 
The characterization of 3D coronary artery motion at different heart rates is essential for the development of valuable data reconstruction strategies in noninvasive coronary angiography. Our study adds to the previous knowledge (5–11) on coronary artery motion the following results: (a) The duration of systole lengthens from 30% at low heart rates (45 beats per minute) to 60% of the cardiac cycle at high heart rates (100 beats per minute). (b) Three-dimensional coronary artery motion velocities vary significantly between arteries and segments, with the highest velocity at the origin of the acute marginal branch of the RCA and the lowest at the origin of the LCX artery. (c) Differences in coronary artery motion velocities between individuals are significantly related to changes in heart rate. (d) Two velocity troughs and two velocity peaks can be identified across different heart rates, with the maximum velocity in early systole and the minimum velocity in mid-diastole. (e) Because of nonproportional shortening of systole and diastole, the two velocity troughs in mid-systole and mid-diastole and the velocity peak in early diastole shift with increasing heart rate to later percentage phases, whereas the early systolic velocity peak remains at a constant interval between 0% and 10%. (f) With increasing heart rate, the minimum mid-diastolic velocity increases, while the width of the mid-diastolic trough successively decreases and eventually disappears. (g) At heart rates greater than 83 beats per minute, the lowest velocities in systole were lower than the lowest velocities in diastole.

Nonproportional shortening of systole and diastole at higher heart rates, with the length of diastole decreasing more than that of systole, is a known physiologic observation (16,17). A similar nonproportional effect also occurs within diastole, where the length of diastasis decreases more than that of the other diastolic phases (ie, isovolumic relaxation, early rapid filling, and late atrial filling) (17). Findings of our study illustrate for the first time, to our knowledge, the effect of these physiologic changes on data reconstruction intervals, by demonstrating a lengthening of the duration of systole from 30% of the cardiac cycle at low heart rates to 60% at high heart rates. This was accompanied by shifts of the mid-systolic and mid-diastolic velocity troughs and shifts of the early diastolic peak. The early systolic peak remained at a constant interval, most probably because the preceding isovolumic contraction phase has been shown to be relatively heart rate independent (21). In line with nonproportional shortening of diastasis, we found the mid-diastolic trough to successively shorten and eventually disappear at heart rates greater than 80 beats per minute. This is in agreement with the results of a study of Chung et al (17), who analyzed the lengths of the diastolic phases and found the duration of diastasis to disappear at a heart rate of approximately 82 beats per minute.

The mechanical consequences of cardiac phases on epicardial arteries are well reflected by coronary artery motion velocities. Velocity peaks were found in early systole in the phase of early rapid ejection and in early diastole in the phase of early rapid filling of the ventricle. The third velocity peak of distal RCA and LCX branches (which did not reach statistical significance when data were averaged) can be assigned to the phase of atrial contraction at end systole and may be explained by the anatomic location of distal RCA and LCX branches in the anterior and posterior atrioventricular grooves. Velocity troughs were found at end systole (20%–30%) in the phase of reduced ejection and in mid-diastole (60%–70%), the latter corresponding to the phase of diastasis that is defined as the phase of near-quiescence of cardiac motion in diastole (21,22). Achenbach et al (6) have considered a velocity trough at 50% of the cardiac cycle to be the "end-systolic trough" in a patient population with a heart rate range of 51–86 beats per minute. Similarly, Hoffmann et al (18) performed "diastolic and systolic" image reconstructions at 50% and 80% of the cardiac cycle in a patient population with a heart rate range of 45–103 beats per minute. Our analysis, which incorporated the duration of systole and diastole, has revealed that the 50% phase of the cardiac cycle can either be a part of systole or a part of diastole, depending on the individual patient's heart rate.

Sixty-four–section CT enables coronary artery imaging with a spatial resolution of 0.4 x 0.4 x 0.4 mm3 combined with a temporal resolution of 83–165 msec (28). Even with this new generation of CT scanners, motion artifacts occur, especially with higher heart rates, and negatively affect image quality and accuracy for diagnosing coronary artery disease (29–32). Wintersperger et al (32) have found the best image quality at 64-section CT coronary angiography to occur in patients with heart rates of less than 65 beats per minute in diastole, and a shift for best image quality at heart rates greater than 75 beats per minute from diastole to systole. In our study, we found a cutoff heart rate of 83 beats per minute at which minimal systolic velocity became smaller than minimal velocities in diastole. On the basis of our findings, CT coronary angiography data should be reconstructed in patients with heart rates of less than 60 beats per minute at 50%–60% of the R-R interval, in patients with heart rates from 60 to 70 beats per minute at 60%–70% of the R-R interval, in patients with heart rates between 71 and 83 beats per minute at 70%–80% of the R-R interval, and in patients with heart rates greater than 83 beats per minute at 30%–40% of the R-R interval. This shift of reconstruction timing also implies individual placement of the center for full tube output when electrocardiographic pulsing is used (33), depending on the actual patient's heart rate.

Our study had limitations. By employing a limited number of landmarks on coronary bifurcations, we have provided descriptions of coronary artery motion only at selected points. However, motion of coronary arteries is nonuniform and is characterized by changes in the magnitude and direction of vessel motion and axial strain (34,35). More than half of our population was receiving ß-blocker medications; therefore, in our study population, actual coronary velocities may have been underestimated because of the lowering of heart rates. We included only patients with no evidence of cardiac disease; however, CT coronary angiography in clinical practice is sometimes performed in patients with impairment of ventricular motion caused by myocardial disease, which would lead to deviations of the motion patterns described above.

We calculated velocities by using the mean heart rate during scanning and not by using the instantaneous heart rate from the beat in question. In addition, the two-segment reconstruction algorithm averages anatomic information from two consecutive heart beats, which may lead to inaccuracies regarding calculation of coronary velocities. We excluded patients with irregular heart rates to minimize these errors. Furthermore, because of the small number of patients with heart rates greater than 83 beats per minute, the upper bound of the 95% confidence interval for determination of the shift of the lowest coronary artery motion velocity from diastole to systole could not be calculated. Closing and opening of the aortic valve may not exactly match the electrophysiologic onset of systole and diastole, potentially missing the phases of "isometric contraction," "protodiastole," and "isometric relaxation" (21). On the other hand, this approach presents the only practicable approach to determining the length of systole and diastole with CT. The fact that the width of the velocity trough decreases and the velocity trough eventually vanishes can be visually appreciated in Figure 4, but, for statistical substantiation of this fact, an arbitrary cutoff had to be chosen. Finally, the absolute temporal resolution used for velocity measurements and for determination of transition from systole to diastole varied from 83 to 165 msec; therefore, the data may misrepresent the precision of these velocity measurements. Nonetheless, we assume that averaging our data across patients and heart rates minimized this mathematical shortcoming.

In conclusion, heart rate substantially affects 3D coronary artery motion velocity through nonproportional shortening of systole and diastole, with consecutive shifts of velocity peaks and troughs. With increasing heart rate, the phase of diastasis progressively disappears, and, at heart rates greater than 83 beats per minute, the smallest coronary velocity translates from diastole to systole. These findings suggest an adaptation of reconstruction time intervals to the individual patient's heart rate and provide the basis for implementation of electrocardiographically controlled tube current modulation at CT coronary angiography.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
References
 


   FOOTNOTES
 

Abbreviations: LAD = left anterior descending • LCX = left circumflex • RCA = right coronary artery • 3D = three-dimensional

Guarantors of integrity of entire study, L.H., S.L., L.D., H.A.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, all authors; clinical studies, L.H., S.L., L.D., T.S., O.G., H.A.; statistical analysis, all authors; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE...
 References
 

  1. Ding Z, Friedman MH. Quantification of 3-D coronary arterial motion using clinical biplane cineangiograms. Int J Card Imaging 2000;16:331–346. [CrossRef][Medline]
  2. Ritchie CJ, Godwin JD, Crawford CR, Stanford W, Anno H, Kim Y. Minimum scan speeds for suppression of motion artifacts in CT. Radiology 1992;185:37–42. [Abstract/Free Full Text]
  3. Kong Y, Morris JJ Jr, McIntosh HD. Assessment of regional myocardial performance from biplane coronary cineangiograms. Am J Cardiol 1971;27:529–537. [CrossRef][Medline]
  4. Potel MJ, Rubin JM, MacKay SA, Aisen AM, Al-Sadir J, Sayre RE. Methods for evaluating cardiac wall motion in three dimensions using bifurcation points of the coronary arterial tree. Invest Radiol 1983;18:47–57. [CrossRef][Medline]
  5. Johnson KR, Patel SJ, Whigham A, Hakim A, Pettigrew RI, Oshinski JN. Three-dimensional, time-resolved motion of the coronary arteries. J Cardiovasc Magn Reson 2004;6:663–673. [CrossRef][Medline]
  6. Achenbach S, Ropers D, Holle J, Muschiol G, Daniel WG, Moshage W. In-plane coronary arterial motion velocity: measurement with electron-beam CT. Radiology 2000;216:457–463. [Abstract/Free Full Text]
  7. Mao S, Lu B, Oudiz RJ, Bakhsheshi H, Liu SC, Budoff MJ. Coronary artery motion in electron beam tomography. J Comput Assist Tomogr 2000;24:253–258. [CrossRef][Medline]
  8. Lu B, Mao SS, Zhuang N, et al. Coronary artery motion during the cardiac cycle and optimal ECG triggering for coronary artery imaging. Invest Radiol 2001;36:250–256. [CrossRef][Medline]
  9. Hofman MB, Wickline SA, Lorenz CH. Quantification of in-plane motion of the coronary arteries during the cardiac cycle: implications for acquisition window duration for MR flow quantification. J Magn Reson Imaging 1998;8:568–576. [Medline]
  10. Saranathan M, Ho VB, Hood MN, Foo TK, Hardy CJ. Adaptive vessel tracking: automated computation of vessel trajectories for improved efficiency in 2D coronary MR angiography. J Magn Reson Imaging 2001;14:368–373. [CrossRef][Medline]
  11. Vembar M, Garcia MJ, Heuscher DJ, et al. A dynamic approach to identifying desired physiological phases for cardiac imaging using multislice spiral CT. Med Phys 2003;30:1683–1693. [CrossRef][Medline]
  12. Vembar M, Walker MJ, Johnson PC. Cardiac imaging using multislice computed tomography scanners: technical considerations. Coron Artery Dis 2006;17:115–123. [CrossRef][Medline]
  13. Manzke R, Kohler T, Nielsen T, Hawkes D, Grass M. Automatic phase determination for retrospectively gated cardiac CT. Med Phys 2004;31:3345–3362. [CrossRef][Medline]
  14. Leschka S, Wildermuth S, Boehm T, et al. Noninvasive coronary angiography with 64-section CT: effect of average heart rate and heart rate variability on image quality. Radiology 2006;241:378–385. [Abstract/Free Full Text]
  15. Leschka S, Husmann L, Desbiolles LM, et al. Optimal image reconstruction intervals for non-invasive coronary angiography with 64-slice CT. Eur Radiol 2006;16:1964–1972. [CrossRef][Medline]
  16. Kovacs SJ Jr. The duration of the QT interval as a function of heart rate: a derivation based on physical principles and a comparison to measured values. Am Heart J 1985;110:872–878. [CrossRef][Medline]
  17. Chung CS, Karamanoglu M, Kovacs SJ. Duration of diastole and its phases as a function of heart rate during supine bicycle exercise. Am J Physiol Heart Circ Physiol 2004;287:H2003–H2008.[Abstract/Free Full Text]
  18. Hoffmann MH, Shi H, Manzke R, et al. Noninvasive coronary angiography with 16-detector row CT: effect of heart rate. Radiology 2005;234:86–97. [Abstract/Free Full Text]
  19. Herzog C, Arning-Erb M, Zangos S, et al. Multi-detector row CT coronary angiography: influence of reconstruction technique and heart rate on image quality. Radiology 2006;238:75–86. [Abstract/Free Full Text]
  20. Flohr T, Ohnesorge B. Heart rate adaptive optimization of spatial and temporal resolution for electrocardiogram-gated multislice spiral CT of the heart. J Comput Assist Tomogr 2001;25:907–923. [CrossRef][Medline]
  21. Wiggers C. Studies of the consecutive phases of the cardiac cycle. I. The duration of the consecutive phases of the cardiac cycle and the criteria for their precise determination. Am J Physiol 1921;56:415–438. [Free Full Text]
  22. Luisada AA, MacCanon DM. The phases of the cardiac cycle. Am Heart J 1972;83:705–711. [CrossRef][Medline]
  23. Baumert B, Plass A, Bettex D, et al. Dynamic cine mode imaging of the normal aortic valve using 16-channel multidetector row computed tomography. Invest Radiol 2005;40:637–647. [CrossRef][Medline]
  24. Austen WG, Edwards JE, Frye RL, et al. A reporting system on patients evaluated for coronary artery disease. Report of the Ad Hoc Committee for Grading of Coronary Artery Disease, Council on Cardiovascular Surgery, American Heart Association. Circulation 1975;51:5–40. [Medline]
  25. Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960;20:37–46. [CrossRef]
  26. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174. [CrossRef][Medline]
  27. Heid IM, Kuchenhoff H, Wellmann J, Gerken M, Kreienbrock L, Wichmann HE. On the potential of measurement error to induce differential bias on odds ratio estimates: an example from radon epidemiology. Stat Med 2002;21:3261–3278. [CrossRef][Medline]
  28. Flohr TG, Stierstorfer K, Ulzheimer S, Bruder H, Primak AN, McCollough CH. Image reconstruction and image quality evaluation for a 64-slice CT scanner with z-flying focal spot. Med Phys 2005;32:2536–2547. [CrossRef][Medline]
  29. Leschka S, Alkadhi H, Plass A, et al. Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J 2005;26:1482–1487. [Abstract/Free Full Text]
  30. Raff GL, Gallagher MJ, O'Neill WW, Goldstein JA. Diagnostic accuracy of noninvasive coronary angiography using 64-slice spiral computed tomography. J Am Coll Cardiol 2005;46:552–557. [Abstract/Free Full Text]
  31. Leber AW, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol 2005;46:147–154. [Abstract/Free Full Text]
  32. Wintersperger BJ, Nikolaou K, von Ziegler F, et al. Image quality, motion artifacts, and reconstruction timing of 64-slice coronary computed tomography angiography with 0.33-second rotation speed. Invest Radiol 2006;41:436–442. [CrossRef][Medline]
  33. Jakobs TF, Becker CR, Ohnesorge B, et al. Multislice helical CT of the heart with retrospective ECG gating: reduction of radiation exposure by ECG-controlled tube current modulation. Eur Radiol 2002;12:1081–1086. [CrossRef][Medline]
  34. Ding Z, Friedman MH. Dynamics of human coronary arterial motion and its potential role in coronary atherogenesis. J Biomech Eng 2000;122:488–492. [CrossRef][Medline]
  35. Duerinckx A, Atkinson DP. Coronary MR angiography during peak-systole: work in progress. J Magn Reson Imaging 1997;7:979–986. [Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
P. Stolzmann, S. Leschka, H. Scheffel, T. Krauss, L. Desbiolles, A. Plass, M. Genoni, T. G. Flohr, S. Wildermuth, B. Marincek, et al.
Dual-Source CT in Step-and-Shoot Mode: Noninvasive Coronary Angiography with Low Radiation Dose
Radiology, October 1, 2008; 249(1): 71 - 80.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
A. C. Weustink, N. R. Mollet, F. Pugliese, W. B. Meijboom, K. Nieman, M. H. Heijenbrok-Kal, T. G. Flohr, L. A. E. Neefjes, F. Cademartiri, P. J. de Feyter, et al.
Optimal Electrocardiographic Pulsing Windows and Heart Rate: Effect on Image Quality and Radiation Exposure at Dual-Source Coronary CT Angiography
Radiology, September 1, 2008; 248(3): 792 - 798.
[Abstract] [Full Text] [PDF]

Home page
 Eur Heart JHome page
L. Husmann, I. Valenta, O. Gaemperli, O. Adda, V. Treyer, C. A. Wyss, P. Veit-Haibach, F. Tatsugami, G. K. von Schulthess, and P. A. Kaufmann
Feasibility of low-dose coronary CT angiography: first experience with prospective ECG-gating
Eur. Heart J., January 2, 2008; 29(2): 191 - 197.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Material
Right arrow All Versions of this Article:
2451061791v1
245/2/567    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Husmann, L.
Right arrow Articles by Alkadhi, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Husmann, L.
Right arrow Articles by Alkadhi, H.

HOME HELP FEEDBACK SUBSCRIPTIONS