# Effect of Cardiac Phase on Cardiac Output Index Derived from Dynamic CT Myocardial Perfusion Imaging

^{1}

^{2}

^{3}

^{4}

^{5}

^{*}

## Abstract

**:**

**Purpose:**The aortic time-enhancement curve obtained from dynamic CT myocardial perfusion imaging can be used to derive the cardiac output (CO) index based on the indicator dilution principle. The objective of this study was to investigate the effect of cardiac phase at which CT myocardial perfusion imaging is triggered on the CO index measurement with this approach.

**Methods:**Electrocardiogram (ECG) gated myocardial perfusion imaging was performed on farm pigs with consecutive cardiac axial scans using a large-coverage CT scanner (Revolution, GE Healthcare) after intravenous contrast administration. Multiple sets of dynamic contrast-enhanced (DCE) cardiac images were reconstructed retrospectively from 30% to 80% R-R intervals with a 5% phase increment. The time-enhancement curve sampled from above the aortic orifice in each DCE image set was fitted with a modified gamma variate function (MGVF). The fitted curve was then normalized to the baseline data point unaffected by the streak artifact emanating from the contrast solution in the right heart chamber. The Stewart–Hamilton equation was used to calculate the CO index based on the integral of the fitted normalized aortic curve, and the results were compared among different cardiac phases.

**Results:**The aortic time-enhancement curves sampled at different cardiac phases were different from each other, especially in the baseline portion of the curve where the effect of streak artifact was prominent. After properly normalizing and denoising with a MGVF, the integrals of the aortic curve were minimally different among cardiac phases (0.228 ± 0.001 Hounsfield Unit × second). The corresponding mean CO index was 4.031 ± 0.028 L/min. There were no statistical differences in either the integral of the aortic curve or CO index among different cardiac phases (p > 0.05 for all phases).

## 1. Introduction

## 2. Methods

#### 2.1. Study Subject

#### 2.2. Dynamic CT Acquisition and Image Reconstruction

#### 2.3. CT Derivation of CO Index

#### 2.3.1. Implementation of Indicator Dilution Method

_{a}(t):

^{−1}).

#### 2.3.2. Sampling Location and Unit Conversion

^{−1}.

#### 2.3.3. Smoothing and Normalization of Sampled Time-Enhancement Curve

_{a}(t) (the denominator in Equation (1)) as the convolution of a transfer function with a probability density function [19,20], while other methods suggested the shape of C

_{a}(t) could be described by a modified gamma variate function (MGVF) [16], as shown in Equation (2):

#### 2.4. Derivation of CO Index at Different Cardiac Phases

#### 2.5. Derivation of CO Index with Ventricular Delineation Method

## 3. Results

#### 3.1. Smoothing and Normalization of Aortic Time-Enhancement Curve Sampled from Dynamic CT Acquisition

#### 3.2. Effect of Cardiac Phase on CO Index Derived from Indicator Dilution Method

^{−1}. There were no statistically significant differences in either the integral of aortic curve or CO index among different cardiac phases (p > 0.05 for all cardiac phases).

#### 3.3. Difference in CO Index Derived from Indicator Dilution and Ventricular Delineation Methods

^{−1}, and the difference with that derived by the indicator dilution method (4.06 ± 0.74 L·min

^{−1}) reached statistical significance (p < 0.05, Figure 6). On average, the CO index estimated with the ventricular delineation method was approximately 73% of that derived with the indicator dilution method. The figure of merit (ratio of standard deviation to mean) of the ventricular delineation method in the 17 CT perfusion studies was also much larger than that of the indicator dilution method (0.304 vs. 0.045), confirming a greater degree of variation in CO assessment when the ventricular delineation method was used.

## 4. Discussion

#### 4.1. Smoothing and Normalization of Sampled Aortic Time-Enhancement Curve

#### 4.2. Effect of Cardiac Phase on CO Index Derived from Indicator Dilution Method

^{−1}) facilitates the concomitant measurement of CO and myocardial perfusion with different dynamic CT myocardial perfusion imaging protocols proposed for assessing myocardial ischemia. Since the mass of tracer injected into each pig was unchanged, the variations in CO measurement among different cardiac phases were contributed entirely by the differences in the sampled aortic time-enhancement curves.

#### 4.3. Difference in CO Index Derived from Indicator Dilution and Ventricular Delineation Methods

^{−1}[8] and 2.45 ± 1.28 L·min

^{−1}[9] vs. 1.07 ± 0.70 L·min

^{−1}). However, our results should be more accurate, given the fact that a more advanced CT scanner was used, which offered a much thinner slice thickness (2.5 mm vs. 8 mm [9] and 10 mm [8]) to optimize the spatial resolution in the transaxial direction; a much faster gantry rotation speed (0.28 sec/rot vs. 0.5 sec/rot [9]), which reduced the overlapping of the left ventricular volume between the end-systole and end-diastole arising from suboptimal temporal resolution; and a larger scan coverage (16 cm vs. 4 cm [8]), which minimized the stepping artifact across the heart due to limited coverage per acquisition.

^{−1}, which was unreasonable for pigs with approximately 60 kg in weight (Figure 6). In contrast, the indicator dilution method was less affected by these limiting factors, leading to a more reasonable mean value of CO index with less fluctuation among the 17 CT perfusion studies.

#### 4.4. Study Limitations

## 5. Conclusion

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## References

- Vincent, J.-L. Understanding cardiac output. Crit. Care
**2008**, 12, 174. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Lang, C.C.; Karlin, P.; Haythe, J.; Lim, T.K.; Mancini, D.M. Peak cardiac power output, measured noninvasively, is a powerful predictor of outcome in chronic heart failure. Circ. Heart Fail.
**2009**, 2, 33–38. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Vignati, C.; Cattadori, G. Measuring Cardiac Output during Cardiopulmonary Exercise Testing. Ann. Am. Thorac Soc.
**2017**, 14 (Suppl. S1), S48–S52. [Google Scholar] [CrossRef] [PubMed] - Klocke, F.J.; Greene, D.G.; Koberstein, R.C. Indicator-dilution measurement of cardiac output with dissolved hydrogen. Circ. Res.
**1968**, 22, 841–853. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Herfkens, R.J.; Axel, L.; Lipton, M.J.; Napel, S.; Berninger, W.; Redington, R. Measurement of cardiac output by computed transmission tomography. Investig. Radiol.
**1982**, 17, 550–553. [Google Scholar] [CrossRef] - Raman, S.V.; Tran, T.; Simonetti, O.P.; Sun, B. Dynamic computed tomography to determine cardiac output in patients with left ventricular assist devices. J. Thorac. Cardiovasc. Surg.
**2009**, 137, 1213–1217. [Google Scholar] [CrossRef] [Green Version] - Pienn, M.; Kovacs, G.; Tscherner, M.; Johnson, T.R.; Kullnig, P.; Stollberger, R.; Olschewski, A.; Olschewski, H.; Bálint, Z. Determination of cardiac output with dynamic contrast-enhanced computed tomography. Int J. Cardiovasc. Imaging
**2013**, 29, 1871–1878. [Google Scholar] [CrossRef] - Ludman, P.F.; Coats, A.J.S.; Poole-Wilson, P.A.; Rees, R.S.O. Measurement accuracy of cardiac output in humans: Indicator-dilution technique versus geometric analysis by ultrafast computed tomography. J. Am. Coll Cardiol.
**1993**, 21, 1482–1489. [Google Scholar] [CrossRef] [Green Version] - Mahnken, A.H.; Klotz, E.; Hennemuth, A.; Jung, B.; Koos, R.; Wildberger, J.E.; Günther, R.W. Measurement of cardiac output from a test-bolus injection in multislice computed tomography. Eur. Radiol.
**2003**, 13, 2498–2504. [Google Scholar] [CrossRef] - Garrett, J.S.; Lanzer, P.; Jaschke, W.; Botvinick, E.; Sievers, R.; Higgins, C.B.; Lipton, M.J. Measurement of cardiac output by cine computed tomography. Am. J. Cardiol.
**1985**, 56, 657–661. [Google Scholar] [CrossRef] - Hsiao, E.M.; Rybicki, F.J.; Steigner, M. CT coronary angiography: 256-slice and 320-detector row scanners. Curr. Cardiol. Rep.
**2010**, 12, 68–75. [Google Scholar] [CrossRef] [Green Version] - Huang, I.-L.; Wu, M.-T.; Hu, C.; Mar, G.-Y.; Lee, T.-Y.; So, A. Quantitative low-dose rest and stress CT myocardial perfusion imaging with a whole-heart coverage scanner improves functional assessment of coronary artery disease. Int. J. Cardiol. Heart Vasc.
**2019**, 24, 100381. [Google Scholar] [CrossRef] - Cademartiri, F.; Seitun, S.; Clemente, A.; La Grutta, L.; Toia, P.; Runza, G.; Midiri, M.; Maffei, E. Myocardial blood flow quantification for evaluation of coronary artery disease by computed tomography. Cardiovasc. Diagn Ther.
**2017**, 7, 129–150. [Google Scholar] [CrossRef] [Green Version] - Abualsaud, A.O.; Lowe, B.S.; Guo, K.; Marelli, A.J.; Kaouache, M.; Guo, L.; Jutras, L.; Martucci, G.; Therrien, J. Cardiac output as a predictor in congenital heart disease: Are we stating the obvious? Int. J. Cardiol.
**2016**, 210, 143–148. [Google Scholar] [CrossRef] - So, A.; Wisenberg, G.; Teefy, P.; Yadegari, A.; Bagur, R.; Hadway, J.; Morrison, L.; MacDonald, A.; Gaskin, D.; Butler, J.; et al. Functional CT assessment of extravascular contrast distribution volume and myocardial perfusion in acute myocardial infarction. Int. J. Cardiol.
**2018**, 266, 15–23. [Google Scholar] [CrossRef] - Stewart, G.N. Researches on the Circulation Time and on the Influences which affect it. J. Physiol.
**1987**, 22, 159–183. [Google Scholar] [CrossRef] - So, A.; Imai, Y.; Nett, B.; Jackson, J.; Hsieh, J.; Wisenberg, G.; Teefy, P.; Yadegari, A.; Islam, A.; Lee, T.-Y. Technical Note: Evaluation of a 160-mm/256-row CT scanner for whole-heart quantitative myocardial perfusion imaging. Med. Phys.
**2016**, 43, 4821–4832. [Google Scholar] [CrossRef] - Wang, J.; Lu, H.; Liang, Z.; Eremina, D.; Zhang, G.; Wang, S.; Chen, J.; Manzione, J. An experimental study on the noise properties of x-ray CT sinogram data in Radon space. Phys. Med. Biol.
**2008**, 53, 3327–3341. [Google Scholar] [CrossRef] - Zierler, K. Indicator dilution methods for measuring blood flow, volume, and other properties of biological systems: A brief history and memoir. Ann. Biomed. Eng.
**2000**, 28, 836–848. [Google Scholar] [CrossRef] - Stephenson, J.L. Theory of Measurement of Blood Flow by Dye Dilution Technique. IRE Trans. Med. Electron.
**1958**, PGME-12, 82–88. [Google Scholar] [CrossRef] - Thompson, H.K.; Starmer, C.F.; Whalen, R.E.; Mcintosh, H.D. Indicator Transit Time Considered as a Gamma Variate. Circ. Res.
**1964**, 14, 502–515. [Google Scholar] [CrossRef] [Green Version] - Davenport, R. The Derivation of the Gamma-Variate Relationship for Tracer Dilution Curves. J. Nucl. Med.
**1983**, 24, 945–948. [Google Scholar] - Band, D.M.; Linton, R.A.F.; O’Brien, T.K.; Jonas, M.M.; Linton, N.W.F. The shape of indicator dilution curves used for cardiac output measurement in man. J. Physiol.
**1997**, 498, 225–229. [Google Scholar] [CrossRef] [Green Version] - Fedorov, A.; Beichel, R.; Kalpathy-Cramer, J.; Finet, J.; Fillion-Robin, J.-C.; Pujol, S.; Bauer, C.; Jennings, D.; Fennessy, F.; Sonka, M.; et al. 3D Slicer as an Image Computing Platform for the Quantitative Imaging Network. Magn. Reson. Imaging
**2012**, 30, 1323–1341. [Google Scholar] [CrossRef] [Green Version] - Brook, R.A.; Chiro, G.D. Principles of computer assisted tomography (CAT) in radiographic and radioisotopic imaging. Phys. Med. Biol.
**1976**, 21, 689–732. [Google Scholar] [CrossRef]

**Figure 1.**(

**a**) The temporal change in aortic enhancement was monitored from dynamic CT imaging after an intravenous bolus injection of iodinated contrast solution. (

**b**) The aortic time-enhancement curve C

_{a}(t) was obtained by fitting the experimental data acquired from CT with a modified gamma variate function (MGVF) followed by baseline subtraction. (

**c**) Schematic illustration of derivation of the CO index from the experimental data and C

_{a}(t).

**Figure 2.**Streak artifacts (yellow arrows) emanating from highly attenuating contrast solution in the right atrium at the (

**a**) end-systolic and (

**b**) end-diastolic phases during the initial stage of the first pass. The artifacts spilled over into the adjacent ascending aorta (pink elliptical region), where the aortic time-enhancement curve was sampled for assessment of the CO index. The two proposed baseline subtraction approaches were evaluated against the reference baseline value measured from the middle section of the left ventricle prior to contrast arrival (

**c**).

**Figure 3.**Segmentation of the left ventricular blood pool in the short axis (

**a**,

**e**), horizontal long axis (

**b**,

**f**), and vertical long axis (

**c**,

**g**) at the end-systolic (top row) and end-diastolic (bottom row) phases for assessment of the CO index with the ventricular delineation method. The relative position of the three orthogonal planes at the end-systolic and end-diastolic phases are illustrated in (

**d**,

**h**), respectively.

**Figure 4.**Two curve normalization approaches for assessment of CO index in a pig study—baseline subtraction using (

**a**) the averaged value of all baseline data points and (

**b**) only the baseline data point unaffected by streak artifacts. The curves shown in (

**a**,

**b**) were acquired from a pig at different cardiac phases, ranging from 30% to 80% of the R-R interval, with each cardiac phase represented by a color illustrated in the color-coded bar adjacent to the graphs. The estimated CO indexes corresponding to the two curve normalization approaches are shown in (

**c**,

**d**), respectively. The CO indexes estimated with the second normalization approach among different cardiac phases were less deviated from each other.

**Figure 5.**(

**a**) Integrals of the fitted normalized aortic time-enhancement curves at different cardiac phases in 17 CT myocardial perfusion studies. Curve fitting was performed by using a modified gamma variate function (MVGF) and the fitted curve was normalized to a single baseline point unaffected by streak artifacts. The crosses in the boxplot represent outliers. Both the integrals and the corresponding CO indexes shown in (

**b**) were not statistically different among different cardiac phases.

**Figure 6.**Comparison in the CO indexes assessed with the ventricular delineation and indicator dilution methods in 17 CT myocardial perfusion studies. The cross in the boxplot represents outliers; * denotes p < 0.05 with respect to the indicator dilution method.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Dempsey, S.C.H.; Lee, T.-Y.; Samani, A.; So, A.
Effect of Cardiac Phase on Cardiac Output Index Derived from Dynamic CT Myocardial Perfusion Imaging. *Tomography* **2022**, *8*, 1129-1140.
https://doi.org/10.3390/tomography8020092

**AMA Style**

Dempsey SCH, Lee T-Y, Samani A, So A.
Effect of Cardiac Phase on Cardiac Output Index Derived from Dynamic CT Myocardial Perfusion Imaging. *Tomography*. 2022; 8(2):1129-1140.
https://doi.org/10.3390/tomography8020092

**Chicago/Turabian Style**

Dempsey, Sergio C. H., Ting-Yim Lee, Abbas Samani, and Aaron So.
2022. "Effect of Cardiac Phase on Cardiac Output Index Derived from Dynamic CT Myocardial Perfusion Imaging" *Tomography* 8, no. 2: 1129-1140.
https://doi.org/10.3390/tomography8020092