Effect of Tube Voltage on Diagnostic Performance of Fractional Flow Reserve Derived From Coronary CT Angiography With Machine Learning: Results From the MACHINE Registry
Abstract
OBJECTIVE. Coronary CT angiography (CCTA)-based methods allow noninvasive estimation of fractional flow reserve (cFFR), recently through use of a machine learning (ML) algorithm (cFFRML). However, attenuation values vary according to the tube voltage used, and it has not been shown whether this significantly affects the diagnostic performance of cFFR and cFFRML. Therefore, the purpose of this study is to retrospectively evaluate the effect of tube voltage on the diagnostic performance of cFFRML.
MATERIALS AND METHODS. A total of 525 coronary vessels in 351 patients identified in the MACHINE consortium registry were evaluated in terms of invasively measured FFR and cFFRML. CCTA examinations were performed with a tube voltage of 80, 100, or 120 kVp. For each tube voltage value, correlation (assessed by Spearman rank correlation coefficient), agreement (evaluated by intraclass correlation coefficient and Bland-Altman plot analysis), and diagnostic performance (based on ROC AUC value, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy) of the cFFRML in terms of detection of significant stenosis were calculated.
RESULTS. For tube voltages of 80, 100, and 120 kVp, the Spearman correlation coefficient for cFFRML in relation to the invasively measured FFR value was ρ = 0.684, ρ = 0.622, and ρ = 0.669, respectively (p < 0.001 for all). The corresponding intraclass correlation coefficient was 0.78, 0.76, and 0.77, respectively (p < 0.001 for all). Sensitivity was 100.0%, 73.5%, and 85.0%, and specificity was 76.2%, 79.0%, and 72.8% for tube voltages of 80, 100, and 120 kVp, respectively. The ROC AUC value was 0.90, 0.82, and 0.80 for 80, 100, and 120 kVp, respectively (p < 0.001 for all).
CONCLUSION. CCTA-derived cFFRML is a robust method, and its performance does not vary significantly between examinations performed using tube voltages of 100 kVp and 120 kVp. However, because of rapid advancements in CT and postprocessing technology, further research is needed.
Recently introduced coronary CT angiography (CCTA)-based methods allow noninvasive estimation of fractional flow reserve (cFFR) through the application of computational fluid dynamics or, for onsite evaluation, machine learning (ML), hence improving the diagnostic yield of CCTA [1, 2] and potentially reducing the number of unnecessary invasive procedures [3]. However, attenuation values vary in association with the tube voltage used [4], and it has not previously been shown whether this significantly affects the diagnostic performance of cFFR estimated with an ML algorithm (cFFRML).
In this multicenter study, we retrospectively evaluated the correlation, agreement, and diagnostic performance of prototype cFFRML software for tube voltages of 80, 100, and 120 kVp in relation to invasively measured FFR. However, because of the small number of examinations performed using a tube voltage of 80 kVp, our main focus was on the difference in the diagnostic performance of cFFRML derived from examinations performed using 100 kVp and 120 kVp, although the results for examinations performed using 80 kVp will be presented where appropriate.
Materials and Methods
The present study was conducted according to the principles of the Declaration of Helsinki and good clinical practice. Approval was obtained from the medical ethics committee at each participating center. Two centers obtained written informed consent from all patients [5, 6]. At the remaining three centers, the need for informed consent was waived because of the retrospective design of the study [7–9].
Patient Information
All data in the present study were extracted from the MACHINE consortium database. The MACHINE consortium is a multicenter collaboration between five institutions in North America, Europe, and Asia that are dedicated to the evaluation of on-site cFFR applications, specifically in reference to ML-based techniques [10]. All participating centers have previously evaluated the diagnostic performance of a first-generation prototype software used on site to estimate cFFR through the application of computational fluid dynamics [5–9]. The patient cohorts, CCTA segmentation data, and cFFR sampling locations from these studies have subsequently been used for the validation of the outwardly similar but ML-based software used in this study [10].
In short, after exclusions (Fig. 1), the database consisted of 351 patients from five centers who had undergone both clinical CCTA and invasive coronary angiography (ICA), with invasive FFR measurements having been performed for 525 coronary arteries (118 right coronary arteries, 235 left main or left anterior descending coronary arteries, and 105 circumflex arteries; sample location data were missing for 67 vessels). These vessels had also been evaluated using prototype cFFRML software (cFFR, version 2.1, Siemens Healthcare; currently not commercially available). Of the 351 patients identified from the database, 334 were examined using a tube voltage of either 100 kVp (174 patients [265 vessels]) or 120 kVp (160 patients [225 vessels]). The remaining 17 patients (35 vessels) were examined using an 80-kVp protocol.
Image and Data Acquisition
All CCTA examinations included in this study were performed using first- or second-generation dual-source CT scanners (Definition or Definition FLASH, Siemens Healthcare). Examinations were performed according to clinical routine, with the use of tube voltages of 80, 100, or 120 kVp. All images were reconstructed using a medium-smooth kernel and a slice thickness of 0.75 mm or less with an increment of 0.4 mm. When needed, β-blockers were administered to patients with high heart rates. Sublingual nitroglycerin was routinely administered at four centers (to 299 patients). Additional details are described elsewhere [5–9].
CCTA segmentation data and cFFR sampling locations from these previous studies [5–9] were transferred and processed locally using prototype cFFRML software (cFFR, version 2.1, Siemens Healthcare; currently not commercially available).
The cFFRML model was trained against a database consisting of 12,000 synthetically generated coronary artery trees, reflecting different anatomic and geometric variations as well as different features of stenosis. A total of 28 input variables were extracted for each 3D geometric variation, to learn their respective relationship to the computational fluid dynamics–based cFFR values [11].
Demographic data and FFR and cFFRML values were retrieved from the MACHINE database.
Statistical Evaluation
Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy (including 95% CIs) were calculated using an online statistical calculator (MedCalc, MedCalc Software). Bland-Altman plots were also constructed using the MedCalc statistical software package (version 18.6). All other statistical calculations were performed using IBM SPSS software (version 24, IBM).
Differences between the 100- and 120-kVp groups in terms of invasively measured FFR, age, body weight, body mass index (BMI; weight in kilograms divided by square of height in meters), heart rate (at CCTA acquisition), dose-length product (DLP), and Agatston score were evaluated using the nonparametric Mann-Whitney U test, because a normal distribution could not be assumed for all parameters and subgroups. Further evaluation of the differences between the 80-kVp subgroup and the 100- and 120-kVp subgroups was also performed using the Mann-Whitney U test.
The diagnostic performance of cFFRML in terms of sensitivity, specificity, NPV, PPV, and accuracy in the detection of significant stenosis (defined as an invasively measured FFR ≤ 0.80) [12, 13] was calculated for all groups and for subgroups based on the three tube voltage levels used for the CCTA examinations (hereafter referred to as the tube voltage subgroups). In addition, the ROC AUC value was also calculated. Correlation between the FFR and the cFFRML was calculated for each tube voltage subgroup by use of the Spearman rank correlation coefficient. Agreement was calculated using the intraclass correlation coefficient for FFR values and the Cohen kappa coefficient for detection of significant stenosis. Bland-Altman plots were constructed for the evaluation of systematic bias.
Results
Demographic Characteristics
On a per-patient basis, the median age for the 351 patients included in the study was 63 years. For 342 of the 351 patients (data were missing for nine patients), the median body weight and BMI were 79 kg and 26.8, respectively. The median heart rate at CCTA was 63 beats/min for 233 patients (data were missing for 118 patients), and the median DLP was 506 for 342 patients (data were missing for nine patients).
On a per-vessel basis, the median (range) invasive FFR and cFFRML for the all groups was 0.84 (0.24–1.00) and 0.82 (0.31–1.00), respectively. Of 525 invasively measured FFR values, 212 (40.4%) indicated significant stenosis. Corresponding values for the tube voltage subgroups were 14 (40.0%), 98 (37.0%), and 100 (44.4%) for the 80-, 100-, and 120-kVp groups, respectively. Additional demographic data for each separate tube voltage setting are presented in Tables 1 and 2.
| Group | Age (y), Mean (Range) | Body Weight (kg) | BMI | Heart Rate (beats/min) | Dose-Length Product (mGy × cm) | Agatston Score |
|---|---|---|---|---|---|---|
| All groups (n = 351) | 63 (56–69) | 79 (49, 140) | 26.8 (24.5, 26.8) | 63 (57, 70) | 506 (324, 779) | 238 (38, 680) |
| 80-kV subgroup (n = 17) | 58 (51–64) | 78 (69, 86) | 25.1 (24.2, 27.1) | 63 (59, 65) | 117 (73,151) | 351 (10, 951) |
| 100-kV subgroup (n = 174) | 64 (56–71) | 77 (70, 86) | 26.3 (24.3, 29.0) | 62 (57, 69) | 440 (292, 572) | 171 (21, 607) |
| 120-kV subgroup (n = 160) | 62 (56–69) | 84 (73, 92) | 27.7 (24., 30.8) | 63 (57, 74) | 731 (468, 1025) | 282 (76, 821) |
Note—Except where otherwise indicated, data are median (25th and 75th percentiles). BMI = body mass index (weight in kilograms divided by the square of height in meters).
| Vessels | Invasively Measured FFR | cFFRML |
|---|---|---|
| All groups (n = 525) | 0.84 (0.74, 0.91) | 0.82 (0.70, 0.90) |
| 80-kV subgroup (n = 35) | 0.85 (0.75, 0.95) | 0.78 (0.67, 0.88) |
| 100-kV subgroup (n = 265) | 0.85 (0.75, 0.92) | 0.83 (0.72, 0.91) |
| 120-kV subgroup (n = 225) | 0.83 (0.74, 0.90) | 0.79 (0.69, 0.89) |
Note—Data are median (25th and 75th percentiles). FFR = fractional flow reserve, cFFRML = coronary CT angiography–based estimation of FFR with machine learning algorithm.
Using the nonparametric Mann-Whitney U test, statistically significant differences between the 100- and 120-kVp groups (on a per-patient basis) were noted in terms of body weight (p = 0.002), BMI (p = 0.007), DLP (p < 0.001), and Agatston (calcium) score (p = 0.021). On a per-vessel basis, significant differences were noted in terms of cFFRML (p = 0.025) but not in terms of invasively measured FFR (p = 0.081).
When the 80-kVp subgroup was compared to the other two subgroups, the only significant differences were observed for age (p = 0.033) and DLP (p < 0.001).
Correlation and Agreement
On a per-vessel basis for all groups, the Spearman rank correlation coefficient for cFFRML in relation to invasive FFR was ρ = 0.645 (p < 0.001). For the 80-, 100-, and 120-kVp tube voltage subgroups, the Spear-man rank correlation coefficient for cFFRML in relation to invasively measured FFR was ρ = 0.684 (p < 0.001), ρ = 0.622 (p < 0.001), and ρ = 0.669 (p < 0.001), respectively.
Agreement between cFFRML and invasively measured FFR in terms of the intra-class correlation coefficient was 0.76 (p < 0.001) for all groups. For the 80-, 100-, and 120-kVp tube voltage subgroups, the intra-class correlation coefficient was 0.78, 0.76, and 0.77, respectively (p < 0.001 for all).
The Cohen kappa coefficient for all groups was 0.56, and it was 0.72, 0.52, and 0.57 for the 80-, 100-, and 120-kVp tube voltage subgroups, respectively (p < 0.001 for all).
Bland-Altman plots for tube voltages of 100 and 120 kVp are presented in Figure 2.
Fig. 2A —Bland-Altman plots for tube voltages of 100 kV and 120 kV.
A, Plots show mean and limits of agreement are similar when tube voltages of 100 kV (A) and 120 kV (B) are used. Systematic bias exists, with CCTA-based estimation of coronary fractional flow reserve with machine learning algorithm (cFFRML) overestimating stenosis grade, compared with invasively measured fractional flow reserve (FFR). There are wide limits of agreement for both tube voltages, with increasing errors for lower FFR and cFFRML values. Dotted line denotes line of equality (FFR – cFFRML = 0); dashed lines, limits of agreement; solid horizontal line, mean difference between FFR and cFFRML values; circles, observed differences between FFR and cFFRML; and error bars, 95% CIs of limits of agreement.
Fig. 2B —Bland-Altman plots for tube voltages of 100 kV and 120 kV.
B, Plots show mean and limits of agreement are similar when tube voltages of 100 kV (A) and 120 kV (B) are used. Systematic bias exists, with CCTA-based estimation of coronary fractional flow reserve with machine learning algorithm (cFFRML) overestimating stenosis grade, compared with invasively measured fractional flow reserve (FFR). There are wide limits of agreement for both tube voltages, with increasing errors for lower FFR and cFFRML values. Dotted line denotes line of equality (FFR – cFFRML = 0); dashed lines, limits of agreement; solid horizontal line, mean difference between FFR and cFFRML values; circles, observed differences between FFR and cFFRML; and error bars, 95% CIs of limits of agreement.
Detection of Significant Stenosis
The sensitivity, specificity, PPV, NPV, and accuracy of cFFRML in terms of detection of significant stenosis are shown in Table 3. The ROC AUC value for all groups was 0.84 (p < 0.001), and it was 0.90, 0.82, and 0.84 for the 80-, 100-, and 120-kVp tube voltage subgroups, respectively (p < 0.001 for all) (Fig. 3).
| Vessels | Sensitivity | Specificity | PPV | NPV | Accuracy |
|---|---|---|---|---|---|
| All groups (n = 525) | 80.7 (74.7–85.8) | 76.4 (71.3–81.0) | 69.8 (65.2–74.0) | 85.4 (81.5–88.5) | 78.1 (74.3–81.6) |
| 80-kV subgroup (n = 35) | 100.0 (76.8–100.0) | 76.2 (52.8–91.8) | 73.7 (56.6–85.7) | 100.0 | 85.7 (69.7–95.2) |
| 100-kV subgroup (n = 265) | 73.5 (63.6–81.9) | 79.0 (72.1–85.0) | 67.3 (60.0–73.9) | 83.5 (78.3–89.7) | 77.0 (71.4–81.9) |
| 120-kV subgroup (n = 225) | 85.0 (76.5–91.4) | 72.8 (64.1–80.4) | 71.4 (65.0–77.1) | 85.9 (79.0–90.7) | 78.2 (72.3–83.4) |
Note—Data are percentage (95% CI). Significant stenosis was defined by an invasively measured fractional flow reserve value of 0.80 or less. PPV = positive predictive value, NPV = negative predictive value.
Fig. 3A —ROC curves for detection of significant stenosis (per-vessel analysis).
A, ROC curves for all vessels (n = 525) (A) and for each subgroup as categorized according to tube voltage level used in coronary CT angiography: 80 kV (n = 35) (B), 100 kV (n = 265) (C), or 120 kV (n = 225) (D). ROC AUC value was 0.84 for all vessels, and 0.90, 0.82, and 0.84 for 80-, 100-, and 120-kV subgroups, respectively (p < 0.001 for all).
Fig. 3B —ROC curves for detection of significant stenosis (per-vessel analysis).
B, ROC curves for all vessels (n = 525) (A) and for each subgroup as categorized according to tube voltage level used in coronary CT angiography: 80 kV (n = 35) (B), 100 kV (n = 265) (C), or 120 kV (n = 225) (D). ROC AUC value was 0.84 for all vessels, and 0.90, 0.82, and 0.84 for 80-, 100-, and 120-kV subgroups, respectively (p < 0.001 for all).
Fig. 3C —ROC curves for detection of significant stenosis (per-vessel analysis).
C, ROC curves for all vessels (n = 525) (A) and for each subgroup as categorized according to tube voltage level used in coronary CT angiography: 80 kV (n = 35) (B), 100 kV (n = 265) (C), or 120 kV (n = 225) (D). ROC AUC value was 0.84 for all vessels, and 0.90, 0.82, and 0.84 for 80-, 100-, and 120-kV subgroups, respectively (p < 0.001 for all).
Fig. 3D —ROC curves for detection of significant stenosis (per-vessel analysis).
D, ROC curves for all vessels (n = 525) (A) and for each subgroup as categorized according to tube voltage level used in coronary CT angiography: 80 kV (n = 35) (B), 100 kV (n = 265) (C), or 120 kV (n = 225) (D). ROC AUC value was 0.84 for all vessels, and 0.90, 0.82, and 0.84 for 80-, 100-, and 120-kV subgroups, respectively (p < 0.001 for all).
Discussion
During the past decade, CCTA has been proved to be a reliable method for detection of coronary artery stenosis, and it is the recommended primary diagnostic tool for patients with a low or intermediate risk of coronary artery disease before testing [14–16]. The method has a high NPV, thus allowing the exclusion of coronary artery disease with a high degree of certainty. However, the PPV and specificity are somewhat less impressive, and further evaluation with ICA is often required when CCTA findings are positive. When ICA is performed, visual evaluation of a perceived stenosis can be further enhanced by functional assessment in the form of measurement of the FFR, the value of which was shown by Tonino et al. [12] in the FAME (FFR versus Angiography Multivessel Evaluation) study. Today, FFR is considered to be the reference standard in stenosis evaluation, and it expresses the relative decrease in coronary artery flow caused by a stenosis during maximal coronary microvascular dilation. Clinically, FFR is determined by measuring the upstream and downstream pressure of a stenosis [17, 18] with use of a dedicated pressure gauge catheter.
However, previous studies have shown that even when preceded by CCTA, approximately 30% of all ICA examinations have negative findings in terms of significant stenosis [19]. ICA and, hence, FFR are invasive and therefore are expensive methods that require the use of additional radiation and contrast agent and also add a small but significant procedural risk for the patient [20, 21]. Thus, the development of robust noninvasive alternatives is indeed needed to improve the selection process and decrease both the monetary cost and the risk of unnecessary invasive procedures. During the past half decade, a number of CCTA-based methods have become available—most notably, cFFR [1, 2] but also measurement of the transluminal contrast attenuation gradient [22] and semiautomated quantification of coronary artery plaque composition [23]. The common characteristic of these methods is that they can be applied to an ordinary CCTA dataset without the requirement for any additional radiation or contrast medium administration (as opposed to, for example, CT perfusion). Because these methods are CT based, they are also all dependent on attenuation values as a basis for segmentation and quantification.
The iodine k-edge is located at approximately 33 keV, which means that a lower peak tube voltage of 70–80 kVp improves iodine-related image contrast compared with protocols using higher tube voltages [24]. This effect is often desirable in CT angiography and has been used to decrease the amount of contrast agent given to elderly patients and patients with impaired renal function [25, 26]. As a drawback, blooming artifacts resulting from calcifications will be more prominent when a lower tube voltage is used [27]. However, standard attenuation values are based on a tube voltage of 120 kVp, and the use of other tube voltages is accompanied by a corresponding deviation in attenuation values [4]. Apart from the variation in plaque composition itself, this could possibly account for some of the large variation in coronary plaque attenuation and characterization that was reported in a previous study [28]. Although this may not necessarily affect the accuracy of manual stenosis evaluation [29], the effect could be more conspicuous when attenuation-based segmentation algorithms are used. Hitherto unpublished data from our center does indeed indicate that the accuracy of some semiautomated plaque quantification software may be severely impaired unless there is compensation for this effect. This also raises a question regarding the extent to which the performance of other quantitative attenuation-dependent, software-based methods, such as cFFRML, are affected by the tube voltage.
In the present study, a total of 334 CCTA examinations were performed with a tube voltage of either 100 or 120 kVp, and a total of 490 invasively measured FFR values with corresponding cFFRML estimates were obtained for the same patients, making the study material uniquely suited for the evaluation of the effect of tube voltage on the accuracy of cFFRML.
The differences in demographic and clinical characteristics between patients in the different tube voltage subgroups were generally small. The 120-kVp subgroup did have a significantly higher Agatston score than the 100-kVp group, but there was no significant difference in the total stenosis burden between the groups, as expressed by the invasive FFR values. Apart from that, significant differences between the 100- and 120-kVp groups were found only in terms of body weight or BMI and dose expressed as DLP; however, this was to be expected because body weight in itself is generally used for triage to protocols with different tube voltages. The 17 patients in the 80-kVp subgroup were slightly (but significantly) younger, but they were otherwise demographically similar to patients in the 100- and 120-kVp subgroups.
All groups combined and the 100- and 120-kVp subgroups yielded very similar results in terms of agreement, correlation, and the ROC AUC value, and no significant differences were observed in terms of sensitivity, specificity, PPV, NPV, or accuracy. Hence, cFFRML seems to be a robust technique with comparable performance, regardless of whether 100 or 120 kVp is used. Of interest, the much smaller 80-kVp subgroup seemed to show even better results in terms of correlation and agreement between cFFRML and invasively measured FFR and also in terms of the ROC AUC value. With such a small group, the reason for this is difficult to identify, but the result could possibly stem from the better iodine-related contrast that can be expected when this tube voltage is used. Also, patients in the 80-kVp subgroup were slightly younger than those in the 100- and 120-kVp subgroups, and one can only speculate whether this might have resulted in better breath-holding and compliance during the CCTA examination. Fourteen of the 17 examinations performed with a tube voltage of 80 kVp were conducted at one center [7], and small local variations in patient selection, examination technique, or reading or postprocessing technique may have also played a role.
Limitations
The present study has a number of limitations. Although it is a multicenter study, the relevant examinations were all performed before the multicenter MACHINE consortium was formed. Even though examination protocols were similar, no agreed-upon protocol was used by all centers. Consequently, small differences in protocol may exist, as well as slight differences in patient selection criteria and postprocessing procedures, depending on local variations. Also, most of the material was retrospectively collected and processed, with only one of five participating centers having performed a prospective study [5].
All examinations in the study were performed using first- and second-generation dual-source CT scanners. Although these scanners still meet the current recommendations for CCTA [30], current state-of-the-art scanners can perform scanning of the coronary arteries with even lower tube voltage, a lower dose, and a lesser amount of contrast medium [31]. Thus, further research in this area is warranted, especially as the possibilities of using advanced postprocessing as a complement to visual evaluation are ever increasing.
Conclusion
CCTA-derived cFFRML is a robust method, and its performance does not vary significantly between 100- and 120-kVp examinations. However, because of rapid advancements in CT and postprocessing technology, further research is needed.
References
1.
Min JK, Leipsic J, Pencina MJ, et al. Diagnostic accuracy of fractional flow reserve from anatomic CT angiography. JAMA 2012; 308:1237–1245
2.
Nørgaard BL, Leipsic J, Gaur S, et al.; NXT Trial Study Group. Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease: the NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps). J Am Coll Cardiol 2014; 63:1145–1155
3.
Douglas PS, De Bruyne B, Pontone G, et al.; PLATFORM Investigators. 1-Year outcomes of FFRCT-guided care in patients with suspected coronary disease: the PLATFORM study. J Am Coll Cardiol 2016; 68:435–445
4.
Cropp RJ, Seslija P, Tso D, Thakur Y. Scanner and kVp dependence of measured CT numbers in the ACR CT phantom. J Appl Clin Med Phys 2013; 14:4417
5.
Kruk M, Wardziak Ł, Demkow M, et al. Workstation-based calculation of CTA-based FFR for intermediate stenosis. JACC Cardiovasc Imaging 2016; 9:690–699
6.
Yang DH, Kim YH, Roh JH, et al. Diagnostic performance of on-site CT-derived fractional flow reserve versus CT perfusion. Eur Heart J Cardiovasc Imaging 2017; 18:432–440
7.
Renker M, Schoepf UJ, Wang R, et al. Comparison of diagnostic value of a novel noninvasive coronary computed tomography angiography method versus standard coronary angiography for assessing fractional flow reserve. Am J Cardiol 2014; 114:1303–1308
8.
Coenen A, Lubbers MM, Kurata A, et al. Fractional flow reserve computed from noninvasive CT angiography data: diagnostic performance of an on-site clinician-operated computational fluid dynamics algorithm. Radiology 2015; 274:674–683
9.
De Geer J, Sandstedt M, Björkholm A, et al. Software-based on-site estimation of fractional flow reserve using standard coronary CT angiography data. Acta Radiol 2016; 57:1186–1192
10.
Coenen A, Kim YH, Kruk M, et al. Diagnostic accuracy of a machine-learning approach to coronary computed tomographic angiography–based fractional flow reserve: result from the MACHINE consortium. Circ Cardiovasc Imaging 2018; 1:e0072171861
11.
Itu L, Rapaka S, Passerini T, et al. A machine-learning approach for computation of fractional flow reserve from coronary computed tomography. J Appl Physiol (1985) 2016; 121:42–52
12.
Tonino PAL, De Bruyne B, Pijls NHJ, et al.; FAME Study Investigators. Fractional flow reserve versus angiography for guiding percutaneous coronary intervention. N Engl J Med 2009; 360:213–224
13.
Patel MR, Calhoon JH, Dehmer GJ, et al. ACC/AATS/AHA/ASE/ASNC/SCAI/SCCT/STS 2017 appropriate use criteria for coronary revascularization in patients with stable ischemic heart disease: a report of the American College of Cardiology Appropriate Use Criteria Task Force, American Association for Thoracic Surgery, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, and Society of Thoracic Surgeons. J Am Coll Cardiol 2017; 69:2212–2241
14.
Taylor AJ, Cerqueira M, Hodgson JM, et al. ACCF/SCCT/ACR/AHA/ASE/ASNC/NASCI/SCAI/SCMR 2010 appropriate use criteria for cardiac computed tomography: a report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, the Society of Cardiovascular Computed Tomography, the American College of Radiology, the American Heart Association, the American Society of Echo-cardiography, the American Society of Nuclear Cardiology, the North American Society for Cardiovascular Imaging, the Society for Cardiovascular Angiography and Interventions, and the Society for Cardiovascular Magnetic Resonance. Circulation 2010; 122:e525–e555
15.
Raff GL, Chinnaiyan KM, Cury RC, et al.; Society of Cardiovascular Computed Tomography Guidelines Committee. SCCT guidelines on the use of coronary computed tomographic angiography for patients presenting with acute chest pain to the emergency department: a report of the Society of Cardiovascular Computed Tomography Guidelines Committee. J Cardiovasc Comput Tomogr 2014; 8:254–271
16.
Wolk MJ, Bailey SR, Doherty JU, et al.; American College of Cardiology Foundation Appropriate Use Criteria Task Force. ACCF/AHA/ASE/ASNC/HFSA/HRS/SCAI/SCCT/SCMR/STS 2013 multimodality appropriate use criteria for the detection and risk assessment of stable ischemic heart disease: a report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, and Society of Thoracic Surgeons. J Am Coll Cardiol 2014; 63:380–406
17.
Pijls NH, van Son JA, Kirkeeide RL, De Bruyne B, Gould KL. Experimental basis of determining maximum coronary, myocardial, and collateral blood flow by pressure measurements for assessing functional stenosis severity before and after percutaneous transluminal coronary angioplasty. Circulation 1993; 87:1354–1367
18.
Pijls NHJ, De Bruyne B, Peels K, et al. Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses. N Engl J Med 1996; 334:1703–1708
19.
Patel MR, Dai D, Hernandez AF, et al. Prevalence and predictors of nonobstructive coronary artery disease identified with coronary angiography in contemporary clinical practice. Am Heart J 2014; 167:846.e2–852.e2
20.
Dehmer GJ, Weaver D, Roe MT, et al. A contemporary view of diagnostic cardiac catheterization and percutaneous coronary intervention in the United States: a report from the CathPCI Registry of the National Cardiovascular Data Registry, 2010 through June 2011. J Am Coll Cardiol 2012; 60:2017–2031
21.
West R, Ellis G, Brooks N; Joint Audit Committee of the British Cardiac Society and Royal College of Physicians of London. Complications of diagnostic cardiac catheterisation: results from a confidential inquiry into cardiac catheter complications. Heart 2006; 92:810–814
22.
Choi JH, Min JK, Labounty TM, et al. Intracoro-nary transluminal attenuation gradient in coronary CT angiography for determining coronary artery stenosis. JACC Cardiovasc Imaging 2011; 4:1149–1157
23.
Boogers MJ, Broersen A, van Velzen JE, et al. Automated quantification of coronary plaque with computed tomography: comparison with intravascular ultrasound using a dedicated registration algorithm for fusion-based quantification. Eur Heart J 2012; 33:1007–1016
24.
Meyer M, Haubenreisser H, Schoepf UJ, et al. Closing in on the K edge: coronary CT angiography at 100, 80, and 70 kV–initial comparison of a second- versus a third-generation dual-source CT system. Radiology 2014; 273:373–382
25.
Viteri-Ramírez G, García-Lallana A, Simón-Yarza I, et al. Low radiation and low-contrast dose pulmonary CT angiography: comparison of 80 kVp/60 ml and 100 kVp/80 ml protocols. Clin Radiol 2012; 67:833–839
26.
Qi L, Wu SY, Meinel FG, et al. Prospectively ECG-triggered high-pitch 80 kVp coronary computed tomography angiography with 30 mL of 270 mg I/mL contrast material and iterative reconstruction. Acta Radiol 2016; 57:287–294
27.
Kalisz K, Buethe J, Saboo SS, Abbara S, Halliburton S, Rajiah P. Artifacts at cardiac CT: physics and solutions. RadioGraphics 2016; 36:2064–2083
28.
Achenbach S, Raggi P. Imaging of coronary atherosclerosis by computed tomography. Eur Heart J 2010; 31:1442–1448
29.
Saremi F, Achenbach S. Coronary plaque characterization using CT. AJR 2015; 204:[web]W249–W260
30.
Abbara S, Blanke P, Maroules CD, et al. SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography: a report of the Society of Cardiovascular Computed Tomography Guidelines Committee—endorsed by the North American Society for Cardiovascular Imaging (NASCI). J Cardiovasc Comput Tomogr 2016; 10:435–449
31.
Linsen PV, Coenen A, Lubbers MM, Dijkshoorn ML, Ouhlous M, Nieman K. Computed tomography angiography with a 192-slice dual-source computed tomography system: improvements in image quality and radiation dose. J Clin Imaging Sci 2016; 6:44
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: October 10, 2018
Accepted: February 7, 2019
First published: April 30, 2019
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.