ORIGINAL ARTICLE
Diagnostic Accuracy of Coronary CT Angiography: Comparison of Filtered Back Projection and Iterative Reconstruction With Different Strengths Rui Wang, MD, PhD,* U. Joseph Schoepf, MD,† Runze Wu, MD,‡ John W. Nance, Jr, MD,§ Biao Lv, MD,* Hua Yang, MD,∥ Fang Li, MD,* Dongxu Lu,* and Zhaoqi Zhang, MD* Purpose: To investigate the diagnostic accuracy of coronary computed tomographic (CT) angiography (CCTA) using filtered back projection (FBP) and sinogram-affirmed iterative reconstruction (SAFIRE) of different strength factors with invasive coronary angiography as the reference standard. Materials and Methods: Fifty consecutive patients (32 men and 18 women) prospectively underwent electrocardiogram-triggered CCTA on a dual-source CT system. The acquisition window was set depending on the heart rate (HR): HR of less than 60 beats per minute (bpm) at the 70% RR interval, 61 to 80 bpm at 30% to 80% RR interval, and greater than 80 bpm at 30% to 50% RR interval; 100 kV and 359 to 377 mA s for patients with a body mass index of less than 24 kg/m2, and 410 to 438 mA s at 120 kV for patients with a body mass index of 24 kg/m2 or greater. Image data were reconstructed using both FBP and SAFIRE. Sinogram-affirmed iterative reconstruction series were reconstructed using 3 different strength factors. Two blinded observers independently assessed the image quality and image impression of each coronary segment using a 4-point scale (1, non-diagnostic; and 4, excellent). Image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured. Filtered back projection and all SAFIRE series were independently evaluated for coronary artery stenosis (>50%), and their diagnostic accuracy was compared with invasive coronary angiography. Results: Statistically significant increases in SNR and CNR were obtained when higher strength factors were used. The highest SNR and CNR were found with the highest SAFIRE strength factor of 5; however, this strength also resulted in a more unfamiliar, “plasticlike” image appearance. Imaging quality scores of FBP and different SAFIRE strengths were 3.37 ± 0.49, 3.41 ± 0.47, 3.52 ± 0.30, and 3.48 ± 0.35, respectively (P < 0.001). The diagnostic accuracies were 92.91%, 93.76%, 95.28%, and 94.94% on per-segment level, respectively (P = 0.993). A tendency toward higher diagnostic performance was observed with SAFIRE strength factor 3 on per-segment analysis, albeit without reaching statistical significance. The effective radiation dose equivalent was 5.7 ± 1.6 mSv. Conclusion: Sinogram-affirmed iterative reconstruction provides significant improvements in image noise, SNR, and CNR compared with FBP, which are progressive with increasing SAFIRE strength factors. Sinogram-affirmed iterative reconstruction strength factor 3 or 5 is recommended for use with CCTA. From the *Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, Beijing, China; †Department of Radiology and Radiological Science, Medical University of South Carolina, Charleston, SC; ‡Siemens Healthcare China, Beijing, China; §The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Hospital, Baltimore, MD; and ∥Affiliated Hospital, Hebei United University, Hebei, China. Received for publication May 22, 2013; accepted August 30, 2013. Reprints: Zhaoqi Zhang, MD, Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 100029 Beijing, China (e‐mail: [email protected]). Conflict of interest statement: UJS is a consultant for and receives research support from Bayer, Bracco, GE, Medrad, and Siemens; and Runze Wu is an employee of Siemens Healthcare. The other authors declare no conflict of interest. Copyright © 2014 by Lippincott Williams & Wilkins
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
Key Words: computed tomography (CT), coronary arteries, radiation dose, image reconstruction (J Comput Assist Tomogr 2014;38: 179–184)
I
terative reconstruction techniques are becoming increasingly available for routine clinical use on many commercial computed tomographic (CT) systems, largely owing to growth in computational power. It has been demonstrated that compared with filtered back projection (FBP), the application of iterative reconstruction on coronary CT angiography (CCTA) provides superior image quality and potential for radiation dose reduction in both unselected patients1 and patients with obesity2 or heavy calcifications.3 However, the iterative reconstruction process results in subjective alterations in image noise patterns and high-frequency border appearance compared to FBP,4 and the resulting unfamiliar image impression (sometimes referred to as “plasticlike”) may negatively affect diagnostic confidence. Vendors have provided several adjustable parameters to try to address this problem, such as blending algorithms with FBP and selectable iterative reconstruction strength factors or levels. Theoretically, these factors can be manipulated to suit user preference. However, using different parameters not only affects the subjective image impression but also influences the image noise and sharpness of vessel borders, potentially affecting the diagnostic accuracy of examinations. Therefore, the objective of this study was to evaluate the effect of using different strength factors of sonogram-affirmed iterative reconstruction (SAFIRE) on subjective and objective image quality and diagnostic accuracy in CCTA using invasive coronary angiography (ICA) as the reference standard.
MATERIALS AND METHODS Patients From April to July 2012, 50 patients with suspected coronary artery disease (CAD) scheduled for ICA were prospectively enrolled in the present study. Exclusion criteria were previous reaction to iodinated contrast media, heart failure (New York Heart Association class III or IV), arrhythmias (atrial fibrillation, etc.), renal insufficiency (serum creatinine >1.4 mg/dL), and prior coronary artery bypass grafts and/or stents. The study was approved by the institutional ethics committee. Written informed consent was obtained from each subject.
CCTA Acquisition Technique and Image Reconstruction All CCTA examinations were performed on a secondgeneration dual-source CT system (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany) using a prospectively electrocardiogram (ECG)-triggered sequential image acquisition protocol. The tube potential was selected based on the body mass index (BMI) of patients, and the tube current was modulated by automatic exposure control: 100 kV and 359 to 377 mA s for www.jcat.org
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
179
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
Wang et al
patients with a BMI of less than 24 kg/m2, and 410 to 438 mA s at 120 kV for patients with a BMI of 24 kg/m2 or more. The padding of the acquisition window was set depending on the heart rate (HR): HR less than 60 beats per minute (bpm) at the 70% RR interval, 61 to 80 bpm at a 30% to 80% RR interval, and greater than 80 bpm at a 30% to 50% RR interval. The examination range was from the tracheal bifurcation to the diaphragm, and the studies were acquired in a craniocaudal direction. Bolus tracking was performed with a region of interest (ROI) placed within the root of the ascending aorta, and image acquisition was automatically triggered 6 seconds after intra-aortic attenuation reached the predefined threshold of 100 Hounsfield units. The collimation was 2 64 0.6 mm; gantry rotation time was 280 milliseconds (ms). No β-blockers were given before the image study. Fifty to 60 mL of contrast agent (iopromide, 370-mgI/mL; Ultravist, Bayer Healthcare, Berlin, Germany) was intravenously injected by a dual-head power injector (Stellant D, Medrad, Indianola, Pa) through an 18-gauge needle placed in the right antecubital vein with injection rates of 5 to 5.5 mL/s. The CCTA studies were reconstructed using both FBP and SAFIRE algorithms. Image reconstruction was performed at the cardiac phase with the least motion with the following parameters: section thickness, 0.75 mm; reconstruction increment, 0.5 mm; and temporal resolution, 75 ms. Filtered back projection series were reconstructed with the “B26f ” vascular kernel; SAFIRE series used the corresponding “I26f ” reconstruction kernel. Sinogram-affirmed iterative reconstruction series were
reconstructed using strength factors 1, 3, or 5, with the strength factor indicating the number of iteration loops.5
Image Quality Analysis CCTA Interpretation All image series were transferred to a commercial workstation (MultiModality Workplace [MMWP], Siemens) for image analysis. Coronary CT angiography image series were evaluated by 2 observers, with 5 and 13 years of experience in cardiovascular imaging, respectively. Evaluations were conducted over 4 weeks to minimize the observers’ recall bias between different reconstructions of the same patient. Interobserver disagreements were resolved by a third experienced observer as an adjudicator. The mean CT attenuation, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were used as objective image quality parameters. Image noise (standard deviation [SD] of ROI measurement) was measured by manually prescribing ROIs at consistent locations within the aortic root, left ventricular chamber, and subcutaneous fat. Calcification or plaque within the aortic wall was carefully avoided during ROI placement. For all measurements, the size, shape, and position of ROIs were kept constant (Fig. 1). The SNR and the CNR were calculated with the following formula: SNR = mean_lumen / SD_lumen and CNR = (mean_lumen − mean_fat) / SD_fat, respectively, where mean_ lumen is the mean CT value within the aorta and left ventricular chamber, mean_fat is the mean CT value of the subcutaneous fat, and SD is the standard deviation (image noise) within the ROI.
FIGURE 1. Image noise measurements at the level of the aortic root in CCTA series reconstructed with FBP and SAFIRE with different strength factors. Image noise, expressed as the standard deviation of Hounsfield unit measurements was 25.7 Hounsfield units (HU) for FBP (A), 22.8 HU for SAFIRE strength factor 1 (B), 17 HU for SAFIRE strength factor 3 (C), and 11.3 HU for SAFIRE strength factor 5 (D).
180
www.jcat.org
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
TABLE 1. Patients’ Characteristics and Radiation Dose Characteristic
Data
Age, yrs Sex, M/F HR, bpm BMI, kg/m2 Diabetes Hypertension Dyslipidemia Smokers DLP, mGy Effective radiation dose equivalent, mSv
62.5 ± 5.3 32/18 66.3 ± 3.1 26.3 ± 3.9 37 (74%) 41 (82%) 25 (50%) 32 (64%) 407.14 ± 114.29 5.7 ± 1.6
Image quality of coronary segments was scored using a 4-point scale as previously described6: 1, poor; 2, fair; 3, good; and 4, excellent, with the score based on vessel opacification, structural discontinuity, artifacts, and image noise. Scores of 2 to 4 were considered as diagnostic image quality. To evaluate the subjective impression of overall image presentation, we derived a score for plasticlike appearance of reconstruction series using a 4-point scale: a score of 1 indicated a lack of plasticlike appearance; 2 indicated minimal plasticlike appearance; 3 indicated moderate plasticlike appearance; and 4 indicated marked plasticlike appearance. Coronary CT angiography interpretation was performed on the axial reconstructions, maximum intensity projections, and curved multiplanar reformations. Axial sections and maximum intensity projections were used to identify coronary lesions. Curved multiplanar reformations were used to quantify severity of stenosis. The coronary artery segments were evaluated using the American Heart Association 15-segment model.7 All segments with a minimal diameter of 1.5 mm at their origin were included. Each segment was classified as normal (smooth wall or tapering border), as nonsignificantly stenotic (luminal irregularities or lumen diameter narrowing
Diagnostic Accuracy of Coronary CT Angiography: Comparison of Filtered Back Projection and Iterative Reconstruction With Different Strengths Rui Wang, MD, PhD,* U. Joseph Schoepf, MD,† Runze Wu, MD,‡ John W. Nance, Jr, MD,§ Biao Lv, MD,* Hua Yang, MD,∥ Fang Li, MD,* Dongxu Lu,* and Zhaoqi Zhang, MD* Purpose: To investigate the diagnostic accuracy of coronary computed tomographic (CT) angiography (CCTA) using filtered back projection (FBP) and sinogram-affirmed iterative reconstruction (SAFIRE) of different strength factors with invasive coronary angiography as the reference standard. Materials and Methods: Fifty consecutive patients (32 men and 18 women) prospectively underwent electrocardiogram-triggered CCTA on a dual-source CT system. The acquisition window was set depending on the heart rate (HR): HR of less than 60 beats per minute (bpm) at the 70% RR interval, 61 to 80 bpm at 30% to 80% RR interval, and greater than 80 bpm at 30% to 50% RR interval; 100 kV and 359 to 377 mA s for patients with a body mass index of less than 24 kg/m2, and 410 to 438 mA s at 120 kV for patients with a body mass index of 24 kg/m2 or greater. Image data were reconstructed using both FBP and SAFIRE. Sinogram-affirmed iterative reconstruction series were reconstructed using 3 different strength factors. Two blinded observers independently assessed the image quality and image impression of each coronary segment using a 4-point scale (1, non-diagnostic; and 4, excellent). Image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured. Filtered back projection and all SAFIRE series were independently evaluated for coronary artery stenosis (>50%), and their diagnostic accuracy was compared with invasive coronary angiography. Results: Statistically significant increases in SNR and CNR were obtained when higher strength factors were used. The highest SNR and CNR were found with the highest SAFIRE strength factor of 5; however, this strength also resulted in a more unfamiliar, “plasticlike” image appearance. Imaging quality scores of FBP and different SAFIRE strengths were 3.37 ± 0.49, 3.41 ± 0.47, 3.52 ± 0.30, and 3.48 ± 0.35, respectively (P < 0.001). The diagnostic accuracies were 92.91%, 93.76%, 95.28%, and 94.94% on per-segment level, respectively (P = 0.993). A tendency toward higher diagnostic performance was observed with SAFIRE strength factor 3 on per-segment analysis, albeit without reaching statistical significance. The effective radiation dose equivalent was 5.7 ± 1.6 mSv. Conclusion: Sinogram-affirmed iterative reconstruction provides significant improvements in image noise, SNR, and CNR compared with FBP, which are progressive with increasing SAFIRE strength factors. Sinogram-affirmed iterative reconstruction strength factor 3 or 5 is recommended for use with CCTA. From the *Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, Beijing, China; †Department of Radiology and Radiological Science, Medical University of South Carolina, Charleston, SC; ‡Siemens Healthcare China, Beijing, China; §The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Hospital, Baltimore, MD; and ∥Affiliated Hospital, Hebei United University, Hebei, China. Received for publication May 22, 2013; accepted August 30, 2013. Reprints: Zhaoqi Zhang, MD, Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 100029 Beijing, China (e‐mail: [email protected]). Conflict of interest statement: UJS is a consultant for and receives research support from Bayer, Bracco, GE, Medrad, and Siemens; and Runze Wu is an employee of Siemens Healthcare. The other authors declare no conflict of interest. Copyright © 2014 by Lippincott Williams & Wilkins
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
Key Words: computed tomography (CT), coronary arteries, radiation dose, image reconstruction (J Comput Assist Tomogr 2014;38: 179–184)
I
terative reconstruction techniques are becoming increasingly available for routine clinical use on many commercial computed tomographic (CT) systems, largely owing to growth in computational power. It has been demonstrated that compared with filtered back projection (FBP), the application of iterative reconstruction on coronary CT angiography (CCTA) provides superior image quality and potential for radiation dose reduction in both unselected patients1 and patients with obesity2 or heavy calcifications.3 However, the iterative reconstruction process results in subjective alterations in image noise patterns and high-frequency border appearance compared to FBP,4 and the resulting unfamiliar image impression (sometimes referred to as “plasticlike”) may negatively affect diagnostic confidence. Vendors have provided several adjustable parameters to try to address this problem, such as blending algorithms with FBP and selectable iterative reconstruction strength factors or levels. Theoretically, these factors can be manipulated to suit user preference. However, using different parameters not only affects the subjective image impression but also influences the image noise and sharpness of vessel borders, potentially affecting the diagnostic accuracy of examinations. Therefore, the objective of this study was to evaluate the effect of using different strength factors of sonogram-affirmed iterative reconstruction (SAFIRE) on subjective and objective image quality and diagnostic accuracy in CCTA using invasive coronary angiography (ICA) as the reference standard.
MATERIALS AND METHODS Patients From April to July 2012, 50 patients with suspected coronary artery disease (CAD) scheduled for ICA were prospectively enrolled in the present study. Exclusion criteria were previous reaction to iodinated contrast media, heart failure (New York Heart Association class III or IV), arrhythmias (atrial fibrillation, etc.), renal insufficiency (serum creatinine >1.4 mg/dL), and prior coronary artery bypass grafts and/or stents. The study was approved by the institutional ethics committee. Written informed consent was obtained from each subject.
CCTA Acquisition Technique and Image Reconstruction All CCTA examinations were performed on a secondgeneration dual-source CT system (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany) using a prospectively electrocardiogram (ECG)-triggered sequential image acquisition protocol. The tube potential was selected based on the body mass index (BMI) of patients, and the tube current was modulated by automatic exposure control: 100 kV and 359 to 377 mA s for www.jcat.org
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
179
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
Wang et al
patients with a BMI of less than 24 kg/m2, and 410 to 438 mA s at 120 kV for patients with a BMI of 24 kg/m2 or more. The padding of the acquisition window was set depending on the heart rate (HR): HR less than 60 beats per minute (bpm) at the 70% RR interval, 61 to 80 bpm at a 30% to 80% RR interval, and greater than 80 bpm at a 30% to 50% RR interval. The examination range was from the tracheal bifurcation to the diaphragm, and the studies were acquired in a craniocaudal direction. Bolus tracking was performed with a region of interest (ROI) placed within the root of the ascending aorta, and image acquisition was automatically triggered 6 seconds after intra-aortic attenuation reached the predefined threshold of 100 Hounsfield units. The collimation was 2 64 0.6 mm; gantry rotation time was 280 milliseconds (ms). No β-blockers were given before the image study. Fifty to 60 mL of contrast agent (iopromide, 370-mgI/mL; Ultravist, Bayer Healthcare, Berlin, Germany) was intravenously injected by a dual-head power injector (Stellant D, Medrad, Indianola, Pa) through an 18-gauge needle placed in the right antecubital vein with injection rates of 5 to 5.5 mL/s. The CCTA studies were reconstructed using both FBP and SAFIRE algorithms. Image reconstruction was performed at the cardiac phase with the least motion with the following parameters: section thickness, 0.75 mm; reconstruction increment, 0.5 mm; and temporal resolution, 75 ms. Filtered back projection series were reconstructed with the “B26f ” vascular kernel; SAFIRE series used the corresponding “I26f ” reconstruction kernel. Sinogram-affirmed iterative reconstruction series were
reconstructed using strength factors 1, 3, or 5, with the strength factor indicating the number of iteration loops.5
Image Quality Analysis CCTA Interpretation All image series were transferred to a commercial workstation (MultiModality Workplace [MMWP], Siemens) for image analysis. Coronary CT angiography image series were evaluated by 2 observers, with 5 and 13 years of experience in cardiovascular imaging, respectively. Evaluations were conducted over 4 weeks to minimize the observers’ recall bias between different reconstructions of the same patient. Interobserver disagreements were resolved by a third experienced observer as an adjudicator. The mean CT attenuation, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were used as objective image quality parameters. Image noise (standard deviation [SD] of ROI measurement) was measured by manually prescribing ROIs at consistent locations within the aortic root, left ventricular chamber, and subcutaneous fat. Calcification or plaque within the aortic wall was carefully avoided during ROI placement. For all measurements, the size, shape, and position of ROIs were kept constant (Fig. 1). The SNR and the CNR were calculated with the following formula: SNR = mean_lumen / SD_lumen and CNR = (mean_lumen − mean_fat) / SD_fat, respectively, where mean_ lumen is the mean CT value within the aorta and left ventricular chamber, mean_fat is the mean CT value of the subcutaneous fat, and SD is the standard deviation (image noise) within the ROI.
FIGURE 1. Image noise measurements at the level of the aortic root in CCTA series reconstructed with FBP and SAFIRE with different strength factors. Image noise, expressed as the standard deviation of Hounsfield unit measurements was 25.7 Hounsfield units (HU) for FBP (A), 22.8 HU for SAFIRE strength factor 1 (B), 17 HU for SAFIRE strength factor 3 (C), and 11.3 HU for SAFIRE strength factor 5 (D).
180
www.jcat.org
© 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Comput Assist Tomogr • Volume 38, Number 2, March/April 2014
TABLE 1. Patients’ Characteristics and Radiation Dose Characteristic
Data
Age, yrs Sex, M/F HR, bpm BMI, kg/m2 Diabetes Hypertension Dyslipidemia Smokers DLP, mGy Effective radiation dose equivalent, mSv
62.5 ± 5.3 32/18 66.3 ± 3.1 26.3 ± 3.9 37 (74%) 41 (82%) 25 (50%) 32 (64%) 407.14 ± 114.29 5.7 ± 1.6
Image quality of coronary segments was scored using a 4-point scale as previously described6: 1, poor; 2, fair; 3, good; and 4, excellent, with the score based on vessel opacification, structural discontinuity, artifacts, and image noise. Scores of 2 to 4 were considered as diagnostic image quality. To evaluate the subjective impression of overall image presentation, we derived a score for plasticlike appearance of reconstruction series using a 4-point scale: a score of 1 indicated a lack of plasticlike appearance; 2 indicated minimal plasticlike appearance; 3 indicated moderate plasticlike appearance; and 4 indicated marked plasticlike appearance. Coronary CT angiography interpretation was performed on the axial reconstructions, maximum intensity projections, and curved multiplanar reformations. Axial sections and maximum intensity projections were used to identify coronary lesions. Curved multiplanar reformations were used to quantify severity of stenosis. The coronary artery segments were evaluated using the American Heart Association 15-segment model.7 All segments with a minimal diameter of 1.5 mm at their origin were included. Each segment was classified as normal (smooth wall or tapering border), as nonsignificantly stenotic (luminal irregularities or lumen diameter narrowing
