HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yamamuro, M. Articles by Konishi, J. Search for Related Content PubMed PubMed Citation Articles by Yamamuro, M. Articles by Konishi, J.

Cardiac Functional Analysis with Multi–Detector Row CT and Segmental Reconstruction Algorithm: Comparison with Echocardiography, SPECT, and MR Imaging¹

¹ From the Departments of Nuclear Medicine and Diagnostic Imaging (M.Y., E.T., S.K., H.T., J.K.), Cardiovascular Surgery (T.N., M.K.), and Cardiovascular Medicine (M.O., R.H., T. Kimura, T. Kita), Kyoto University Graduate School of Medicine, 54 Shogoinkawahara, Sakyo-ku, Kyoto 606-8507, Japan; and Department of Nuclear Medicine, Hokkaido University Graduate School of Medicine, Sapporo, Japan (N.T.). Received August 10, 2003; revision requested October 22; final revision received March 20, 2004; accepted May 28. Address correspondence to M.Y. (e-mail: myamamu@kuhp.kyoto-u.ac.jp).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To evaluate accuracy of cardiac functional analysis with multi–detector row computed tomography (CT) and segmental reconstruction algorithm over a range of heart rates.

MATERIALS AND METHODS: Institutional review board approval was obtained. Informed consent was not required. Multi–detector row CT (500-msec rotation time, 8 x 1-mm detector collimation) and magnetic resonance (MR) imaging were performed in 50 patients (28 men, 22 women; age range, 46–84 years; mean age, 67 years). Two-dimensional echocardiography was performed in 41 patients, and electrocardiographically (ECG)-gated single photon emission computed tomography (SPECT) was performed in 27. End-diastolic volume (EDV), end-systolic volume (ESV), ejection fraction (EF), and left ventricular (LV) mass were estimated with multi–detector row CT and compared with values estimated with MR imaging, which served as the reference standard. Additionally, EF values estimated with multi–detector row CT, echocardiography, and SPECT were compared with those estimated with MR imaging. Systemic error and degree of agreement of global functional parameters measured with MR imaging and other modalities were assessed. In a second analysis, linear regression analysis was added.

RESULTS: EF estimated with multi–detector row CT agreed and correlated well with EF estimated with MR imaging (bias ± standard deviation, –1.2% ± 4.6; r = 0.96). Agreement and correlation were similar for EDV (–0.35 mL ± 15.2; r = 0.97), ESV (1.1 mL ± 8.6; r = 0.99), and LV mass (2.5 mL ± 15.0; r = 0.96). Standard deviation of EF difference between multi–detector row CT and MR imaging was significantly less than that between echocardiography and MR imaging (P < .001) or that between SPECT and MR imaging (P < .001).

CONCLUSION: Various LV functional parameters were measured with multi–detector row CT with a segmental approach, and measurements correlated and agreed with those obtained with MR imaging. Moreover, functional analysis with multi–detector row CT was more accurate than that with two-dimensional echocardiography or ECG-gated SPECT.

© RSNA, 2005

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Evaluation of global left ventricular (LV) function, especially ejection fraction (EF), and mass is important in the treatment of various cardiac diseases. To date, cardiac functional assessment has been performed with various noninvasive modalities, such as echocardiography (1,2), nuclear medicine (3), single–detector row helical computed tomography (CT) (4), electron beam CT (5–8), and magnetic resonance (MR) imaging (9–19). Cardiac MR imaging provides excellent temporal and spatial resolution, allows image acquisitionin any desired plane, and has a high degree of accuracy and reproducibility concerning quantitative measurements. In addition, MR imaging can be used to measure LV volume, without assumptions about LV cavity geometry. Thus, MR imaging is currently considered the reference standard in assessment of cardiac function (1,3,9–12,18,19).

In the past few years, multi–detector row CT has been increasingly used for noninvasive coronary artery imaging (20–31). In the evaluation of cardiac function, multi–detector row CT with a temporal resolution of 125–250 msec has been shown to be promising by comparing with biplanar cineventriculography (32) or MR imaging (33). However, it was indicated that reconstructed images obtained in patients with a high heart rate were of low quality because of motion artifacts; thus, manual tracing of endocardial contours had limited accuracy (32,33). A segmental reconstruction algorithm that uses data from several heartbeats has been introduced to further improve temporal resolution (34). The basic principle of segmental reconstruction is that data needed to reconstruct one image are collected retrospectively from several heartbeats by dividing into several segments. Each segment obtained in one cardiac cycle represents a shorter time period. Thus, a segmental reconstruction algorithm is considered to be effective in shortening the temporal resolution and reducing motion artifacts in patients with a high heart rate. Thus, the aim of this study was to evaluate the accuracy of cardiac functional analysis with multi–detector row CT and a segmental reconstruction algorithm over a range of heart rates.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Phantom Studies
To evaluate artifacts of reconstruction images obtained with two retrospective electrocardiographically (ECG)-gated reconstruction methods, phantom (AZ-631N; Anzai Medical, Tokyo, Japan) studies were performed with various heart rates (every 5% of cardiac cycle, from 40 to 100 beats per minute). Method 1 involved a half reconstruction algorithm, and method 2 involved a segmental reconstruction algorithm that used data from several heartbeats. CT examinations were performed by using a multi–detector row CT scanner with eight detector rows (Aquilion Multi V1.10 JR 002 system; Toshiba, Otawara, Japan). In the scanning protocol, a collimation of 8 x 1 mm, a helical pitch of 2.0, and a rotation time of 500 msec were used. The temporal resolution with a half approach was 250 msec, while the temporal resolution with a segmental approach was further improved, depending on heart rate (Fig 1). Each helical CT scan was obtained with a tube voltage of 135 kV and a tube current of 200 mA. Image reconstruction was performed with a 1-mm increment by using both algorithms for each of these heart rates (35,36). Twenty sets of reconstructions at every 5% of the cardiac cycle, ranging from 0% to 95%, were performed in each study. One short-axis image and two long-axis cine images were reoriented from these sets of reconstruction data by using cardiac MPR software (Toshiba). The optimal end-diastolic and end-systolic phases were visually determined with these cine images.

View larger version (25K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Graph shows temporal resolution of the multi-detector row CT system with segmental approach (rotation time, 500 msec; helical pitch, 2.0) at various heart rates. bpm = Beats per minute.

Human Studies
Patients referred for coronary multi–detector row CT from August 2002 to July 2003 for clinical reasons were included in this retrospective study. Among these 50 patients, MR imaging was used to assess cardiac function within 10 days before or after multi–detector row CT, during which time the condition of patients was stable. The study group consisted of 28 men (age range, 47–83 years; mean age, 67 years) and 22 women (age range, 46–84 years; mean age, 67 years). A test for the proportion with normal distribution with a 5% significance level was used to analyze the proportion of men and women, a two-sample t test with a 5% significance level was used to analyze the age difference between men and women, and an F test with a 5% significance level was used to analyze equality of variance between the age of men and women. No statistically significant difference was observed regarding age or sex. Mean heart rate during acquisition of CT scans ranged from 49 to 106 beats per minute (mean ± standard deviation, 71 beats per minute ± 13). A total of 20 patients had aortic and/or mitral valve disease, 12 had myocardial infarction, 12 had angina pectoris, two had infectious endocarditis, two had sarcoidosis, one had pericarditis, and one had dilated cardiomyopathy. During the same period, conventional two-dimensional echocardiography was performed in 41 patients, and ECG-gated single photon emission computed tomography (SPECT) was performed in 27. Patient characteristics and final diagnoses are shown in Table 1. Our institutional review board approved this retrospective study. Informed consent was not required.

CT images were analyzed by an experienced observer (S.K.) without any clinical information. Manual adjustments of endocardial and epicardial borders of each short-axis image were performed. As previously described, papillary muscles were regarded as being part of the LV cavity (10,18). Subsequently, end-diastolic volume (EDV), end-systolic volume (ESV) (both were measured in milliliters), and LV mass (measured in grams) were calculated on the basis of the Simpson rule. LV mass was calculated as a product of the specific gravity of the myocardium (ie, 1.05 g/cm³) and integrated LV myocardial area (10). Finally, the percentage of EF was calculated with EDV and ESV data (38).

Interobserver Variability of CT Measurements
Because the repeatability of multi–detector row CT measurements is relevant to the amount of agreement, interobserver variability was tested by comparing measurements obtained by two experienced observers (S.K., H.T.) at different times and without any clinical information by using data sets obtained in the first 20 patients.

MR Imaging and Image Interpretation
MR imaging was performed with a 1.5-T whole-body imager (Symphony; Siemens, Erlangen, Germany), with multiple surface coils connected to phased-array receivers. Breath-hold cine MR imaging was performed with the segmented steady-state free precession sequence (13–15). Imaging parameters were as follows: repetition time msec/echo time msec, 3.6/1.8; flip angle, 70°; seven to 15 lines per segment; matrix, 208 x 256; and field of view, 340 x 340. Cine MR images of 10–12 contiguous sections with 10-mm section thickness were obtained in the short-axis plane, covering the entire LV from the base to the apex, to acquire three-dimensional LV data (18,19).

MR images were analyzed by an observer (E.T., with 10 years of experience) without any clinical information but with the aid of commercially available software (Argus; Siemens). Image analysis was followed with manual correction of the LV border, as with multi–detector row CT. Subsequently, various functional parameters were measured with the same method.

Echocardiography and Image Interpretation
For two-dimensional echocardiography, patients underwent imaging in the left lateral decubitus position by using a commercially available system (Sonos 5500; Hewlett-Packard, Palo Alto, Calif). Images were obtained by an observer (M.O., with 8 years of experience) using a 3.5-MHz transducer in the parasternal (long- and short-axis) and apical (two- and four-chamber views) planes and were saved in cine loop format. EF was calculated by the same observer, who was blinded to other information obtained with two- and four-chamber images, and who used the previously validated biplanar Simpson rule (1,2,38–41).

ECG-gated SPECT and Image Interpretation
Each patient underwent exercise thallium 201 (²⁰¹Tl) SPECT, as described previously (18,42). At the peak exercise level, approximately 74 MBq of ²⁰¹Tl was injected intravenously. Approximately 10 minutes after the termination of exercise, initial SPECT was performed. Approximately 3–4 hours after termination of the initial examination, an additional 37 MBq of ²⁰¹Tl was injected, and reinjection SPECT was performed. ECG-gated SPECT images were obtained with a dual-head gamma camera (Millennium; GE Medical Systems, Milwaukee, Wis) equipped with low-energy thin section collimation (30 projections over 180°, eight frames per cardiac cycle, 60 seconds per projection). Two energy windows were used (ie, 30% windows centered on the 70- and 167-keV peaks). ECG-gated perfusion SPECT images were prefiltered with Butterworth filter (order, five; voxel size, 7.2 mm; cutoff frequency, 0.40 cycles per pixel) (18). A zoom factor of 1.28 was used. Data were reconstructed by using a filtered back-projection technique with no attenuation or scatter correction. The EF was automatically calculated with reinjection SPECT images by using the Germano algorithm (3,18,19,43).

Statistical Analysis
Data are expressed as mean ± standard deviation. Evaluation of agreement between artifact scores of the phantom was performed by using statistics. Systemic error and the degree of agreement of global functional parameters obtained with multi–detector row CT and MR imaging were assessed according to the method described by Bland and Altman (44). Systemic error and the degree of agreement of EF, as calculated with echocardiography and MR imaging data and ECG-gated SPECT and MR imaging data, respectively, were assessed with the same method. The degree of agreement between the two methods was determined as the mean difference (bias), standard deviation of the differences, limits of agreement (mean ± 2 standard deviations), standard error of the mean difference, and 95% confidence interval of the mean difference. A one-sample t test at the 5% significance level was used to determine whether the resulting difference from zero, as an under- or overestimation with multi–detector row CT, was significant. F ratios were used to describe the equality of the standard deviation of difference of EF values obtained with multi–detector row CT with a segmental reconstruction algorithm and MR imaging to EF values obtained with echocardiography and MR imaging and EF values obtained with ECG-gated SPECT and MR imaging.

In a second analysis, linear regression analysis was used to compare the functional values obtained with multi–detector row CT, echocardiography, and ECG-gated SPECT with MR imaging. F ratios were used to describe the equality of standard error of estimates of EF between the correlation of multi–detector row CT with a segmental reconstruction algorithm and MR imaging, between echocardiography and MR imaging, and between ECG-gated SPECT and MR imaging. A P value of less than .05 was considered to indicate statistical significance.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Phantom Studies
The results of phantom studies are shown in Table 2. The value for evaluation of agreement between the blurring artifact scores was 0.58, while that between the stair-step artifacts was 0.73. Various artifacts were observed in patients with higher heart rates (65 beats per minute) who were evaluated with a half reconstruction algorithm; no substantial artifacts were found in patients evaluated with a segmental reconstruction algorithm, except in those with a heart rate of 80 beats per minute. In patients with higher heart rates (65 beats per minute), artifacts of reconstructed images obtained with a half reconstruction algorithm were generally so severe that determination of the myocardial outline was difficult (Fig 2).

View larger version (98K): [in this window] [in a new window] [Download PPT slide]   Figure 2a. End-diastolic and end-systolic vertical long-axis CT scans of a phantom were reconstructed with half and segmental reconstruction algorithms at a heart rate of 95 beats per minute. (a) End-diastolic and (b) end-systolic vertical long-axis CT scans obtained with half approach. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans obtained with segmental approach. Severe artifacts were observed with a half approach, especially in the end-systolic phase, while no severe artifacts were observed with a segmental approach. Artifacts of reconstructed images obtained with a half approach were so severe that determination of the myocardial outline was difficult.

View larger version (101K): [in this window] [in a new window] [Download PPT slide]   Figure 2b. End-diastolic and end-systolic vertical long-axis CT scans of a phantom were reconstructed with half and segmental reconstruction algorithms at a heart rate of 95 beats per minute. (a) End-diastolic and (b) end-systolic vertical long-axis CT scans obtained with half approach. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans obtained with segmental approach. Severe artifacts were observed with a half approach, especially in the end-systolic phase, while no severe artifacts were observed with a segmental approach. Artifacts of reconstructed images obtained with a half approach were so severe that determination of the myocardial outline was difficult.

View larger version (90K): [in this window] [in a new window] [Download PPT slide]   Figure 2c. End-diastolic and end-systolic vertical long-axis CT scans of a phantom were reconstructed with half and segmental reconstruction algorithms at a heart rate of 95 beats per minute. (a) End-diastolic and (b) end-systolic vertical long-axis CT scans obtained with half approach. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans obtained with segmental approach. Severe artifacts were observed with a half approach, especially in the end-systolic phase, while no severe artifacts were observed with a segmental approach. Artifacts of reconstructed images obtained with a half approach were so severe that determination of the myocardial outline was difficult.

View larger version (88K): [in this window] [in a new window] [Download PPT slide]   Figure 2d. End-diastolic and end-systolic vertical long-axis CT scans of a phantom were reconstructed with half and segmental reconstruction algorithms at a heart rate of 95 beats per minute. (a) End-diastolic and (b) end-systolic vertical long-axis CT scans obtained with half approach. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans obtained with segmental approach. Severe artifacts were observed with a half approach, especially in the end-systolic phase, while no severe artifacts were observed with a segmental approach. Artifacts of reconstructed images obtained with a half approach were so severe that determination of the myocardial outline was difficult.

View larger version (149K): [in this window] [in a new window] [Download PPT slide]   Figure 3a. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (148K): [in this window] [in a new window] [Download PPT slide]   Figure 3b. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (113K): [in this window] [in a new window] [Download PPT slide]   Figure 3c. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 3d. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (111K): [in this window] [in a new window] [Download PPT slide]   Figure 3e. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (123K): [in this window] [in a new window] [Download PPT slide]   Figure 3f. End-diastolic and end-systolic short-axis and vertical long-axis CT scans were reconstructed with a segmental reconstruction algorithm in a patient with a heart rate of 98 beats per minute. (a) Typical end-diastolic and (b) corresponding end-systolic short-axis CT scans. (c) End-diastolic and (d) end-systolic vertical long-axis CT scans. Epicardial (black line in a and b) and endocardial (white line in a and b) tracings exclude epicardial fat and include papillary muscles within ventricular cavity. No severe artifact was found on these scans. End-diastolic and end-systolic vertical long-axis scans were also reconstructed with a half reconstruction algorithm in the same patient. (e) End-diastolic and (f) end-systolic vertical long-axis CT scans. Artifacts of reconstructed images obtained with a half reconstruction algorithm were more severe than those obtained with a segmental approach. Of note, the end systolic size of the LV cavity appeared larger on scans obtained with a half approach than on scans obtained with a segmental approach. Angina pectoris was diagnosed in this patient.

View larger version (31K): [in this window] [in a new window] [Download PPT slide]   Figure 4a. Graphs created with Bland-Altman analysis of (a) LV EF, (b) EDV, (c) ESV, and (d) LV mass, as measured with multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging in 50 patients. Graphs show close agreement between multi-detector row CT and MR imaging. The slopes of the regression lines were not significantly different from 0. No significant differences of the mean difference from 0 were found.

View larger version (36K): [in this window] [in a new window] [Download PPT slide]   Figure 4b. Graphs created with Bland-Altman analysis of (a) LV EF, (b) EDV, (c) ESV, and (d) LV mass, as measured with multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging in 50 patients. Graphs show close agreement between multi-detector row CT and MR imaging. The slopes of the regression lines were not significantly different from 0. No significant differences of the mean difference from 0 were found.

View larger version (32K): [in this window] [in a new window] [Download PPT slide]   Figure 4c. Graphs created with Bland-Altman analysis of (a) LV EF, (b) EDV, (c) ESV, and (d) LV mass, as measured with multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging in 50 patients. Graphs show close agreement between multi-detector row CT and MR imaging. The slopes of the regression lines were not significantly different from 0. No significant differences of the mean difference from 0 were found.

View larger version (34K): [in this window] [in a new window] [Download PPT slide]   Figure 4d. Graphs created with Bland-Altman analysis of (a) LV EF, (b) EDV, (c) ESV, and (d) LV mass, as measured with multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging in 50 patients. Graphs show close agreement between multi-detector row CT and MR imaging. The slopes of the regression lines were not significantly different from 0. No significant differences of the mean difference from 0 were found.

View larger version (27K): [in this window] [in a new window] [Download PPT slide]   Figure 5a. Graphs created with Bland-Altman analysis of LV EF in 41 patients, as measured with (a) multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging or (b) echocardiography and MR imaging. Graphs clearly show that the dispersion between multi-detector row CT and MR imaging was significantly less than that between echocardiography and MR imaging (P < .001).

View larger version (29K): [in this window] [in a new window] [Download PPT slide]   Figure 5b. Graphs created with Bland-Altman analysis of LV EF in 41 patients, as measured with (a) multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging or (b) echocardiography and MR imaging. Graphs clearly show that the dispersion between multi-detector row CT and MR imaging was significantly less than that between echocardiography and MR imaging (P < .001).

View larger version (28K): [in this window] [in a new window] [Download PPT slide]   Figure 6a. Graphs created with Bland-Altman analysis of LV EF in 27 patients, as measured with (a) multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging or (b) ECG-gated SPECT and MR imaging. Graphs clearly show that the dispersion between multi-detector row CT and MR imaging was significantly less than that between ECG-gated SPECT and MR imaging (P < .001).

View larger version (30K): [in this window] [in a new window] [Download PPT slide]   Figure 6b. Graphs created with Bland-Altman analysis of LV EF in 27 patients, as measured with (a) multi-detector row CT (MDCT) with a segmental reconstruction algorithm and MR imaging or (b) ECG-gated SPECT and MR imaging. Graphs clearly show that the dispersion between multi-detector row CT and MR imaging was significantly less than that between ECG-gated SPECT and MR imaging (P < .001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Our results demonstrate that the LV values, including EF, EDV, ESV, and LV mass, obtained with multi–detector row CT and a segmental reconstruction algorithm correlated and agreed with those obtained with MR imaging over a wide range of heart rates. Moreover, functional analysis with multi–detector row CT and a segmental reconstruction algorithm was more accurate than analysis performed with two-dimensional echocardiography or ECG-gated SPECT.

Among the many factors that may affect the accuracy of LV functional measurements with multi–detector row CT, the main limitation is the fact that temporal resolution is worse with multi–detector row CT than with MR imaging (approximately 50 msec). A single helical CT study with a temporal resolution of 400 msec (4) and recent multi–detector row CT studies with a temporal resolution of 125–250 msec performed with a half or biphasic reconstruction algorithm showed close correlation with cineventriculography (32) and MR imaging (33) studies, in terms of EF. Because of the limited temporal resolution, however, systolic images, especially those obtained in patients with a higher heart rate, were shown to be of lower quality (32,33). Additionally, ESV was overestimated, while EDV was not significantly different (4). As a result, underestimation of EF was noted in these studies. In our study, artifacts of reconstructed images obtained with a half approach (temporal resolution of 250 msec) were generally severe if the heart rate was high. In addition, overestimation of ESV and underestimation of EF were observed when a half approach was applied. Juergens et al (32) suggested that data reconstruction algorithms that use segmental data from several heartbeats will likely allow optimization of the analysis of cardiac function with multi–detector row CT. In fact, our phantom experiments demonstrated that a segmental reconstruction algorithm, in which several sets of heartbeat data were used, was more appropriate than a half reconstruction algorithm in terms of reducing motion artifacts, especially in patients with high heart rates.

In the current human studies, no substantial motion artifact was observed, even in patients with a high heart rate, when a segmental reconstruction approach was used. In addition, the various LV functional values obtained with multi–detector row CT and a segmental reconstruction method were shown to correlate and agree with those obtained with MR imaging, which was in agreement with the findings of the preliminary study of Halliburton et al (34). Of note, the standard deviation of EF difference between multi–detector row CT with a segmental approach and MR imaging was significantly less than that between echocardiography and MR imaging or ECG-gated SPECT and MR imaging. Thus, functional analysis with multi–detector row CT and a segmental approach was considered more accurate than functional analysis with two-dimensional echocardiography or ECG-gated SPECT. Previous reports (1,2) suggest that two-dimensional echocardiography is a poor modality to use in the assessment of LV volume and function when ventricular geometry is not uniform. ECG-gated SPECT is also limited when large LV perfusion defects exist (45) or the heart is small (46).

In general, a segmental reconstruction algorithm is effective in shortening the temporal resolution and reducing motion artifacts. At a certain heart rate, however, temporal resolution is not improved, even with a segmental reconstruction approach. For example, in patients with a heart rate of 80 beats per minute, a temporal resolution of 250 msec is not improved—even with a segmental reconstruction algorithm—which results in increased artifacts in our phantom study. Thus, the functional assessment of certain heart rates may still be less than ideal. A more rapid rotation time (approximately 0.4 seconds per rotation) has been attained with multi–detector row CT (23,34), which will make it possible to shorten and stabilize the temporal resolution with a segmental approach. As a result, functional assessment with multi–detector row CT will be further improved.

Our results reveal good repeatability among multi–detector row CT measurements, especially with a segmental approach, which indicates that determination of LV border with multi–detector row CT and a segmental approach is accurate and reliable. In fact, we did not encounter substantial difficulties due to motion artifacts when determining the LV border, even in patients with a high heart rate. Although there was no significant difference between the mean value of various functional parameters as determined with multi–detector row CT and MR imaging, the Bland-Altman plot reveals wide limits of agreement. In addition, temporal resolution of a segmental reconstruction algorithm is dependent on heart rate, which is not the case with MR imaging. These findings suggest that LV functional measurements obtained with multi–detector row CT and MR imaging may not be interchangeable.

Recent reports show that the use of a ß-blocker is effective in lowering heart rates and reducing motion artifacts (31). According to the literature, however, a ß-blocker decreases heart rates by an average of only 8 beats per minute (31). In addition, Vogl et al (22) reported that use of a ß-blocker did not produce any effect in some patients. Thus, use of a ß-blocker is not always effective in the reduction of motion artifacts. In addition, premedication is troublesome in clinical settings, because careful observation and a prolonged stay are required (23); therefore, ß-blockers were not routinely used in our institution. Instead, a segmental reconstruction algorithm has been used to improve temporal resolution when the heart rate is high. A segmental approach may not be useful in the precise assessment of coronary arteries, however, especially when the R-R interval is irregular. At present, there is no definitive solution when the heart rate is high and irregular (23). In addition, the use of a ß-blocker changes the LV function (47). Thus, functional parameters, with the exception of LV mass, can be unreliable after such premedication. To obtain functional data as added information, use of a ß-blocker is unacceptable.

Radiation dose reduction is clinically important. A large helical pitch, reduced tube current, increased number of detector rows, and faster rotation time can be used to reduce the radiation dose. For example, a larger helical pitch will lead to reduced radiation exposure. When the helical pitch is greater, however, the temporal resolution with a segmental approach becomes worse, which may decrease data fidelity not only for multi–detector row CT coronary information but also for functional analysis. Reduced tube current would serve to directly reduce radiation exposure. Reduced tube current will, however, cause increased image noise. For these reasons, a larger helical pitch or reduced tube current should be avoided (30). Recently, a multi–detector row CT scanner equipped with more detector rows and a faster rotation time has been introduced, resulting in reduction of the radiation dose. In addition, lowering the tube current during unneeded phases of the cycle (48) is effective for radiation dose reduction. Unfortunately, this strategy may influence the reliability of functional analysis.

The use of multi–detector row CT solely in the assessment of cardiac function does not appear reasonable because of the radiation exposure. It is important, however, that a functional analysis can be conducted by using data obtained with noninvasive coronary imaging without an increase in cost, volume of contrast agent administered, or radiation dose. We believe that multi–detector row CT will be useful in the comprehensive evaluation of coronary arteries and resting LV function.

A limitation of the F test may be the possible violation of the independence assumption, since the variances are about measures taken on the same subject. The critical F values were used for comparison purposes, and statistical significance was not necessarily implied.

In conclusion, the various LV functional parameters of multi–detector row CT with a segmental reconstruction algorithm obtained by using data from several heartbeats correlated and agreed with those obtained with MR imaging. Moreover, the functional analysis with multi–detector row CT with a segmental reconstruction algorithm was more accurate than that with two-dimensional echocardiography or ECG-gated SPECT.

	FOOTNOTES

 
Abbreviations: ECG = electrocardiographic, EDV = end-diastolic volume, EF = ejection fraction, ESV = end-systolic volume, LV = left ventricle

Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, M.Y.; study concepts and design, M.Y.; literature research, M.Y.; clinical studies, E.T., S.K., M.O., R.H., T.N.; experimental studies, M.Y., E.T., S.K., H.T.; data acquisition, E.T., S.K., H.T., M.O., R.H., T.N.; data analysis/interpretation, E.T., S.K., H.T., M.O., R.H.; statistical analysis, M.Y.; manuscript preparation, definition of intellectual content, and editing, M.Y., E.T.; manuscript revision/review, M.O., R.H., T.N., T. Kimura, N.T., M.K., T. Kita, J.K.; manuscript final version approval, all authors

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Buck T, Hunold P, Wentz KU, Tkalec W, Nesser HJ, Erbel R. Tomographic three-dimensional echocardiographic determination of chamber size and systolic function in patients with left ventricular aneurysm. Circulation 1997; 96:4286-4297.[Abstract/Free Full Text]
2. Qin JX, Jones M, Shiota T, et al. Validation of real-time three-dimensional echocardiography for quantifying left ventricular volumes in the presence of a left ventricular aneurysm: in vitro and in vivo studies. J Am Coll Cardiol 2000; 36:900-907.[Abstract/Free Full Text]
3. Bavelaar-Croon CD, Kayser HW, van der Wall EE, et al. Left ventricular function: correlation of quantitative gated SPECT and MR imaging over a wide range of values. Radiology 2000; 217:572-575.[Abstract/Free Full Text]
4. Mochizuki T, Murase K, Higashino H, et al. Two- and three-dimensional CT ventriculography: a new application of helical CT. AJR Am J Roentgenol 2000; 174:203-208.[Abstract/Free Full Text]
5. Lessick J, Sideman S, Azhari H, Marcus M, Grenadier E, Beyar R. Regional three-dimensional geometry and function of left ventricles with fibrous aneurysms: a cine-computed tomography study. Circulation 1991; 84:1072-1086.[Abstract/Free Full Text]
6. Lipton MJ, Higgins CB, Farmer D, Boyd DP. Cardiac imaging with a high-speed cine-CT scanner: preliminary results. Radiology 1984; 152:579-582.[Abstract/Free Full Text]
7. Lipton MJ, Farmer DW, Killebrew EJ, et al. Regional myocardial dysfunction: evaluation of patients with prior myocardial infarction with fast CT. Radiology 1985; 157:735-740.[Abstract/Free Full Text]
8. Rees MR, Feiring AJ, Rumberger JA, MacMillan RM, Clark DL. Heart evaluation by cine CT: use of two new oblique views. Radiology 1986; 159:804-806.[Abstract/Free Full Text]
9. Higgins CB. Which standard has the gold? (editorial). J Am Coll Cardiol 1992; 19:1608-1609.[Medline]
10. Pattynama PM, Lamb HJ, van der Velde EA, van der Wall EE, de Roos A. Left ventricular measurements with cine and spin-echo MR imaging: a study of reproducibility with variance component analysis. Radiology 1993; 187:261-268.[Abstract/Free Full Text]
11. Bellenger NG, Burgess MI, Ray SG, et al. Comparison of left ventricular ejection fraction and volumes in heart failure by echocardiography, radionuclide ventricography, cardiovascular magnetic resonance; are they interchangeable? Eur Heart J 2000; 21:1387-1396.[Abstract/Free Full Text]
12. de Roos A, Kunz P, Lamb H, et al. Magnetic resonance imaging of ischemic heart disease: why cardiac magnetic resonance imaging will play a significant role in the management of patients with coronary artery disease. J Comput Assist Tomogr 1999; 23(suppl 1):S135-S141.
13. Barkhausen J, Ruehm SG, Goyen M, et al. MR evaluation of ventricular function: true fast imaging with steady-state precession versus fast low-angle shot cine MR imaging: feasibility study. Radiology 2001; 219:264-269.[Abstract/Free Full Text]
14. Carr JC, Simonetti O, Bundy J, Li D, Pereles S, Finn JP. Cine MR angiography of the heart with segmented true fast imaging with steady-state precession. Radiology 2001; 219:828-834.[Abstract/Free Full Text]
15. Miller S, Simonetti OP, Carr J, Kramer U, Finn JP. MR imaging of the heart with cine true fast imaging with steady-state precession: influence of spatial and temporal resolutions on left ventricular functional parameters. Radiology 2002; 223:263-269.[Abstract/Free Full Text]
16. Heusch A, Koch JA, Krogmann ON, Korbmacher B, Bourgeois M. Volumetric analysis of the right and left ventricle in a porcine heart model: comparison of three-dimensional echocardiography, magnetic resonance imaging and angiography. Eur J Ultrasound 1999; 9:245-255.[CrossRef][Medline]
17. Lorenz CH, Walker ES, Morgan VL, Klein SS, Graham TP, Jr. Normal human right and left ventricular mass, systolic function, and gender differences by cine magnetic resonance imaging. J Cardiovasc Magn Reson 1999; 1:7-21.[Medline]
18. Tadamura E, Kudoh T, Motooka M, et al. Assessment of regional and global left ventricular function by reinjection Tl-201 SPECT and rest Tc-99m sestamibi ECG-gated SPECT. J Am Coll Cardiol 1999; 33:991-997.[Abstract/Free Full Text]
19. Tadamura E, Kudoh T, Motooka M, et al. Use of technetium-99m sestamibi ECG-gated single-photon emission tomography for the evaluation of left ventricular function following coronary artery bypass graft: comparison with three-dimensional magnetic resonance imaging. Eur J Nucl Med 1999; 26:705-712.[CrossRef][Medline]
20. Nieman K, Oudkerk M, Rensing BJ, et al. Coronary angiography with multi-slice computed tomography. Lancet 2001; 357:599-603.[CrossRef][Medline]
21. Achenbach S, Giesler T, Ropers D, et al. Detection of coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically-gated, multislice spiral computed tomography. Circulation 2001; 103:2535-2538.[Abstract/Free Full Text]
22. Vogl TJ, Abolmaali ND, Diebold T, et al. Techniques for the detection of coronary atherosclerosis: multi-detector row CT coronary angiography. Radiology 2002; 223:212-220.[Abstract/Free Full Text]
23. Nieman K, Cademartiri F, Lemos PA, Raaijmarkers R, Pattynama PM, Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2002; 106:2051-2054.[Abstract/Free Full Text]
24. Schroeder S, Kopp AF, Baumbach A, et al. Noninvasive detection and evaluation of atherosclelotic coronary plaques with multislice computed tomography. J Am Coll Cardiol 2001; 37:1430-1435.[Abstract/Free Full Text]
25. Horiguchi J, Nakanishi T, Ito K. Quantification of coronary artery calcium using multidetector CT and a retrospective ECG-gating reconstruction algorism. AJR Am J Roentgenol 2001; 177:1429-1435.[Abstract/Free Full Text]
26. White RD, Setser RM. Integrated approach to evaluating coronary artery disease and ischemic heart disease. Am J Cardiol 2002; 90:49L-55L.[CrossRef][Medline]
27. Hong C, Becker CR, Schoepf UJ, Ohnesorge B, Bruening R, Reiser MF. Coronary artery calcium: absolute quantification in nonenhanced and contrast-enhanced multi–detector row CT studies. Radiology 2002; 223:474-480.[Abstract/Free Full Text]
28. Willmann JK, Weishaupt D, Lachat M, et al. Electrocardiographically gated multi–detector row CT for assessment of valvular morphology calcification in aortic stenosis. Radiology 2002; 225:120-128.[Abstract/Free Full Text]
29. Hong C, Becker CR, Huber A, et al. ECG-gated reconstructed multi–detector row CT coronary angiography: effect of varying trigger delay on imaging quality. Radiology 2001; 220:712-717.[Abstract/Free Full Text]
30. Kopp AF, Schroeder S, Kuettner A, et al. Coronary arteries: retrospectively ECG-gated multi-detector row CT angiography with selective optimization of the image reconstruction window. Radiology 2001; 221:683-688.[Abstract/Free Full Text]
31. Ropers D, Baum U, Pohle K, et al. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003; 107:664-666.[Abstract/Free Full Text]
32. Juergens KU, Grude M, Fallenberg EM, et al. Using ECG-gated multidetector CT to evaluate global left ventricular myocardial function in patients with coronary artery disease. AJR Am J Roentgenol 2002; 179:1545-1550.[Abstract/Free Full Text]
33. Juergens KU, Grude M, Maintz D, et al. Multi–detector row CT of left ventricular function with dedicated analysis software versus MR imaging: initial experience. Radiology 2004; 230:403-410.[Abstract/Free Full Text]
34. Halliburton SS, Boese JM, Flohr TG, Lieber ML, Kuzmiak SA, White RD. Improved volumetric analysis of the left ventricle using cardiac multislice computed tomography (MSCT) with high temporal resolution image reconstruction (abstr). Radiology 2002; 225(P):389.
35. Ota T, Taguchi K. A navigation tool to assist optimal scan protocol of electrocardiogram (ECG)-gated retrospective reconstruction for multislice helical CT (abstr). Radiology 2001; 221(P):690.
36. 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]
37. Takase K, Sawamura Y, Igarashi K, et al. Demonstration of the artery of Adamkiewicz at multidetector row helical CT. Radiology 2002; 223:39-45.[Abstract/Free Full Text]
38. Dirksen MS, Bax JJ, de Roos A, et al. Usefulness of dynamic multislice computed tomography of left ventricular function in unstable angina pectoris and comparison with echocardiography. Am J Cardiol 2002; 90:1157-1160.[CrossRef][Medline]
39. Schiller NB, Acquatella H, Ports TA, et al. Left ventricular volume from paired biplane two-dimensional echocardiography. Circulation 1979; 60:547-555.[Abstract/Free Full Text]
40. Gordon EP, Schnittger I, Fitzgerald PJ, Williams P, Popp RL. Reproducibility of left ventricular volumes by two-dimensional echocardiography. J Am Coll Cardiol 1983; 2:506-513.[Abstract]
41. Otterstad JE, Froeland G, St John SM, Holme I. Accuracy and reproducibility of biplane two-dimensional echocardiographic measurements of left ventricular dimensions and function. Eur Heart J 1997; 18:507-513.[Abstract/Free Full Text]
42. Takahashi N, Tamaki N, Tadamura E, et al. Combined assessment of regional perfusion and wall motion in patients with coronary artery disease with technetium-99m tetrofosmin. J Nucl Cardiol 1994; 1:29-38.[Medline]
43. Germano G, Kiat H, Kavanagh PB, et al. Automatic quantification of ejection fraction from gated myocardial perfusion SPECT. J Nucl Med 1995; 36:2138-2147.[Abstract/Free Full Text]
44. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307-310.[CrossRef][Medline]
45. Manrique A, Faraggi M, Vera P, et al. Tl-201 and Tc-99m MIBI gated SPECT in patients with large perfusion defects and left ventricular dysfunction: comparison with equilibrium radionuclide angiography. J Nucl Med 1999; 40:805-809.[Abstract/Free Full Text]
46. Mazzanti M, Geremano G, Kait H, et al. Identification of severe and extensive coronary artery disease by automatic measurement of transient ischemic dilation of the left ventricle in dual-isotope myocardial perfusion SPECT. J Am Coll Cardiol 1996; 27:1612-1620.[Abstract]
47. Ooi H, Colucci WS. Pharmac