Jaap M. Groen

The influence of heart rate, slice thickness, and calcification density on calcium scores using 64-slice multidetector computed tomography - A systematic phantom study

Objective:The purpose of this study was to investigate the influence of heart rate, slice thickness, and calcification density on absolute value and variability of calcium score using 64-slice multidetector computed tomography (MDCT). Methods and Materials:Three artificial arteries containing each 3 lesions with varying density were scanned using a moving cardiac phantom at rest and at 50 to 110 beats per minute (bpm) at 10-bpm intervals on a 64-slice MDCT. Images were reconstructed at slice thicknesses (increment) of 0.6 (0.4), 0.75 (0.5), 1.5 (1.5), and 3.0 (3.0) mm. The amount of calcium was expressed as an Agatston score, volume score, and equivalent mass. Results:Absolute coronary artery calcium (CAC) scores decreased [average −37% for low density calcification (LDC)] or increased [average +32% for high density calcification (HDC)] at heart rates over 60 bpm depending on slice thickness and scoring method. Thinner slice thicknesses yielded higher CAC scores. Variability of the CAC scores increased with increasing heart rates especially for low density calcifications (8% at rest vs. 50% at 110 bpm). Variability also increased for thicker slices (average 6% for 0.6 mm vs. 18% for 3.0 mm). Variability was lower for HDC compared with LDC (∼5% for HDC vs. 27% for LDC at 70 bpm, averaged over all methods and slice thicknesses). Conclusion:CAC-scoring is strongly influenced by cardiac motion, calcification density, and slice thickness. CAC scores increase for high density calcifications and decrease for low density calcifications at increasing heart rates. Heart rate should be reduced on 64-slice MDCT to obtain a lower degree of variability of CAC-scoring, preferably below 70 bpm. A thinner slice thickness further enhances the reproducibility.

Relationship between chronic obstructive pulmonary disease and subclinical coronary artery disease in long-term smokers

AIMS Cardiovascular conditions are reported to be the most frequent cause of death in patients with chronic obstructive pulmonary disease (COPD). However, it remains unsettled whether severity of COPD per se is associated with coronary artery disease (CAD) independent of traditional cardiovascular risk factors. The aim of this study was to examine the relationship between the presence and severity of COPD and the amount of coronary artery calcium deposit, an indicator of CAD and cardiac risk, in a large population of current and former long-term smokers. METHODS AND RESULTS In this cross-sectional study, long-term smokers without clinically manifested CAD were recruited from the Danish Lung Cancer Screening Trial and classified according to lung function by the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria. Coronary artery calcium deposit as a measure of subclinical CAD and cardiac risk was evaluated with multi detector computed tomography and the Agatston coronary artery calcium score (CACS). Participants were categorized into five CACS risk classification groups according to the CACS. The population (n = 1535) consisted of 41% participants without COPD, 28% with mild, and 31% with moderate-to-severe COPD (n = 46 with severe COPD). In addition to age, male gender, hypertension, hypercholesterolaemia, and continued smoking, COPD according to GOLD classification were independent predictors of a higher CACS risk classification group in multivariable analysis [odds ratio (OR): 1.28 (1.01-1.63) and OR: 1.32 (1.05-1.67), for mild and moderate-to-severe COPD, respectively, compared with no COPD]. CONCLUSION COPD in long-term smokers is independently correlated with the CACS, while COPD severity per se does not show a dose-response relationship.

Initial results on visualization of coronary artery stents at multiple heart rates on a moving heart phantom using 64-MDCT.

Objective: Evaluation of the image quality of coronary artery stents at various heart rates using Multi Detector Computed Tomography (MDCT). Methods: Nine different coronary stents were attached to a moving heart phantom and scanned using a 64-MDCT with a rotation time of 330 milliseconds (ms). The heart rate of the phantom was varied between 0 and 115 beats per minute (bpm). Two independent methods were used to investigate image quality. After reconstruction the average Houndsfield Unit (HU) value in the stent lumen was measured in the longitudinal and the cross-sectional plane. The stent images were then presented to two radiologists. The radiologists were asked to rank the images from good to bad based on lumen visibility and overall image quality. A second ranking was obtained using the CT density values. Finally two rankings were compared. Results: Compared to the value for air, the HU-values measured in the lumen increased by 50 to 700 HU. Average slope value in the longitudinal plane was 1.7 ± 0.6 HU/bpm, and the average slope value in the cross-sectional plane was 1.7 ± 0.8 HU/bpm. This shows increased attenuation with increasing heart rate and thus a negative correlation between image quality and heart rate in both planes for all stents. The ranking acquired from the radiologists resembled the measured results as they also showed a negative correlation between the two variables. Using the results of the CT density measurements an analysis was done on multi-segment reconstruction (MSR). Conclusion: A negative correlation between the heart rate and image quality of coronary stents was found by two independent methods. MSR showed no benefit for image quality in this study.

64 slice MDCT generally underestimates coronary calcium scores as compared to EBT: A phantom study

The objective of our study was the determination of the influence of the sequential and spiral acquisition modes on the concordance and deviation of the calcium score on 64-slice multi-detector computed tomography (MDCT) scanners in comparison to electron beam tomography (EBT) as the gold standard. Our methods and materials were an anthropomorphic cardio CT phantom with different calcium inserts scanned in sequential and spiral acquisition modes on three identical 64-slice MDCT scanners of manufacturer A and on three identical 64-slice MDCT scanners of manufacturer B and on an EBT system. Every scan was repeated 30 times with and 15 times without a small random variation in the phantom position for both sequential and spiral modes. Significant differences were observed between EBT and 64-slice MDCT data for all inserts, both acquisition modes, and both manufacturers of MDCT systems. High regression coefficients (0.90-0.98) were found between the EBT and 64-slice MDCT data for both scoring methods and both systems with high correlation coefficients (R2>0.94). System A showed more significant differences between spiral and sequential mode than system B. Almost no differences were observed in scanners of the same manufacturer for the Agatston score and no differences for the Volume score. The deviations of the Agatston and Volume scores showed regression dependencies approximately equal to the square root of the absolute score. The Agatston and Volume scores obtained with 64-slice MDCT imaging are highly correlated with EBT-obtained scores but are significantly underestimated (-10% to -2%) for both sequential and spiral acquisition modes. System B is more independent of acquisition mode to calcium score than system A. The Volume score shows no intramanufacturer dependency and its use is advocated versus the Agatston score. Using the same cut points for MDCT-based calcium scores as for EBT-based calcium scores can result in classifying individuals into a too low risk category. System information and scanprotocol is therefore needed for every calcium score procedure to ensure a correct clinical interpretation of the obtained calcium score results.

A new approach to the assessment of lumen visibility of coronary artery stent at various heart rates using 64-slice MDCT

Coronary artery stent lumen visibility was assessed as a function of cardiac movement and temporal resolution with an automated objective method using an anthropomorphic moving heart phantom. Nine different coronary stents filled with contrast fluid and surrounded by fat were scanned using 64-slice multi-detector computed tomography (MDCT) at 50–100 beats/min with the moving heart phantom. Image quality was assessed by measuring in-stent CT attenuation and by a dedicated tool in the longitudinal and axial plane. Images were scored by CT attenuation and lumen visibility and compared with theoretical scoring to analyse the effect of multi-segment reconstruction (MSR). An average increase in CT attenuation of 144 ± 59 HU and average diminished lumen visibility of 29 ± 12% was observed at higher heart rates in both planes. A negative correlation between image quality and heart rate was non-significant for the majority of measurements (P > 0.06). No improvement of image quality was observed in using MSR. In conclusion, in-stent CT attenuation increases and lumen visibility decreases at increasing heart rate. Results obtained with the automated tool show similar behaviour compared with attenuation measurements. Cardiac movement during data acquisition causes approximately twice as much blurring compared with the influence of temporal resolution on image quality.

A model for quantitative correction of coronary calcium scores on multidetector, dual source, and electron beam computed tomography for influences of linear motion, calcification density, and temporal resolution: a cardiac phantom study.

PURPOSE The objective of this study is to quantify the influence of linear motion, calcification density, and temporal resolution on coronary calcium determination using multidetector computed tomography (MDCT), dual source CT (DSCT), and electron beam tomography (EBT) and to find a quantitative method which corrects for the influences of these parameters using a linear moving cardiac phantom. METHODS On a robotic arm with artificial arteries with four calcifications of increasing density, a linear movement was applied between 0 and 120 mm/s (step of 10 mm/s). The phantom was scanned five times on 64-slice MDCT, DSCT, and EBT using a standard acquisition protocol. The average Agatston, volume, and mass scores were determined for each velocity, calcification, and scanner. Susceptibility to motion was quantified using a cardiac motion susceptibility (CMS) index. Resemblance to EBT and physical volume and mass was quantified using aΔ index. RESULTS Increasing motion artifacts were observed at increasing velocities on all scanners, with increasing severity from EBT to DSCT to 64-slice MDCT. The calcium score showed a linear dependency on motion from which a correction factor could be derived. This correction factor showed a linear dependency on the mean calcification density with a good fit for all three scoring methods and all three scanners (0.73≤R2≤0.95). The slope and offset of this correction factor showed a linear dependency on temporal resolution with a good fit for all three scoring methods and all three scanners (0.83≤R2≤0.98). CMS was minimal for EBT and increasing values were observed for DSCT and highest values for 64-slice MDCT. CMS was minimal for mass score and increasing values were observed for volume score and highest values for Agatston score. For all densities and scoring methods DSCT showed on average the closest resemblance to EBT calcium scores. When using the correction factor, CMS index decreased on average by 15% and Δ index decreased by 35%. CONCLUSIONS Calcium scores determined on DSCT and 64-slice MDCT are highly susceptible to motion as compared to EBT. The mass score is less susceptible to motion compared to volume and Agatston score. Calcium scores determined on DSCT bear a closer resemblance to EBT obtained calcium scores than 64-slice MDCT. In addition, the calcium score is highly dependent on the average density of individual calcifications and the dependency of the calcium score on motion showed a linear behavior on calcification density. From these relations, a quantitative method could be derived which corrects the measured calcium score for the influence of linear motion, mean calcification density, and temporal resolution.

Feasibility and accuracy of tissue characterization with dual source computed tomography

PURPOSE To evaluate the feasibility and accuracy of a model for tissue characterization with dual source computed tomography (DSCT). METHODS AND MATERIALS A model for tissue characterization in CT was used with a parameterization of linear attenuation coefficients. Sixteen chemical substances with effective atomic numbers between 5.21 and 13.08 and electron densities between 2.20 and 4.12 x10(23) electrons/cm(3) were scanned at energies of 80 and 140 kV on a DSCT. From the reconstructed dual energy data sets, effective atomic numbers and electron densities of the substances were calculated. RESULTS Our presented model using DSCT approximated the effective atomic numbers and effective electron densities of 16 substances very well. The measured effective atomic numbers deviated 3.4 ± 6.8% (R(2) = 0.994) from theoretical effective atomic numbers. In addition, measured effective electron densities deviated -0.6 ± 2.2% (R(2) = 0.999) from theoretical effective electron densities. CONCLUSION Effective atomic numbers and effective electron densities can be determined with a high accuracy with DSCT. Therefore the model can be of potential benefit for clinical applications of quantitative tissue characterization with DSCT.

Calcium score of small coronary calcifications on multidetector computed tomography: results from a static phantom study.

INTRODUCTION Multi detector computed tomography (MDCT) underestimates the coronary calcium score as compared to electron beam tomography (EBT). Therefore clinical risk stratification based on MDCT calcium scoring may be inaccurate. The aim of this study was to assess the feasibility of a new phantom which enables establishment of a calcium scoring protocol for MDCT that yields a calcium score comparable to the EBT values and to the physical mass. MATERIALS AND METHODS A phantom containing 100 small calcifications ranging from 0.5 to 2.0mm was scanned on EBT using a standard coronary calcium protocol. In addition, the phantom was scanned on a 320-row MDCT scanner using different scanning, reconstruction and scoring parameters (tube voltage 80-135 kV, slice thickness 0.5-3.0mm, reconstruction kernel FC11-FC15 and threshold 110-150 HU). The Agatston and mass score of both modalities was compared and the influence of the parameters was assessed. RESULTS On EBT the Agatston and mass scores were between 0 and 20, and 0 and 3mg, respectively. On MDCT the Agatston and mass scores were between 0 and 20, and 0 and 4 mg, respectively. All parameters showed an influence on the calcium score. The Agatston score on MDCT differed 52% between the 80 and 135kV, 65% between 0.5 and 3.0mm and 48% between FC11 and FC15. More calcifications were detected with a lower tube voltage, a smaller slice thickness, a sharper kernel and a lower threshold. Based on these observations an acquisition protocol with a tube voltage of 100 kV and two reconstructions protocols were defined with a FC12 reconstruction kernel; one with a slice thickness of 3.0mm and a one with a slice thickness of 0.5mm. This protocol yielded an Agatston score as close to the EBT as possible, but also a mass score as close to the physical phantom value as possible, respectively. CONCLUSION With the new phantom one acquisition protocol and two reconstruction protocols can be defined which produces Agatston scores comparable to EBT values and to the physical mass.

Threshold adjusted calcium scoring using CT is less susceptible to cardiac motion and more accurate.

The purpose of this paper is to investigate calcium scoring on computed tomography (CT) using an adjusted threshold depending on the maximum Hounsfield value within the calcification (HU(peak)). The volume of 19 calcifications was retrospectively determined on 64-slice multidetector CT and dual source CT (DSCT) at different thresholds and the threshold associated with the physical volume was determined. In addition, approximately 10 000 computer simulations were done simulating the same process for calcifications with mixed density. Using these data a relation between the HU(peak) and the threshold could be established. Hereafter, this relation was assessed by scanning six calcifications in a phantom at 40-110 beats per minute using DSCT. The influence of motion was determined and the measured calcium scores were compared to the physical volumes and mass. A positive linear correlation was found between the scoring threshold and the HU(peak) of the calcifications both for the phantom measurements as for the computer simulations. Using this relation the individual threshold for each calcification could be calculated. Calcium scores of the moving calcifications determined with an adjusted threshold were approximately 30% less susceptible to cardiac motion compared to standard calcium scoring. Furthermore, these scores approximated the physical volume and mass at least 10% better than the standard calcium scores. The threshold in calcium scoring should be adjusted for each individual calcification based on the HU(peak) of the calcification. Calcium scoring using an adjusted threshold is less susceptible to cardiac motion and more accurate compared to the physical values.

Can nontriggered thoracic CT be used for coronary artery calcium scoring? A phantom study.

PURPOSE Coronary artery calcium score, traditionally based on electrocardiography (ECG)-triggered computed tomography (CT), predicts cardiovascular risk. However, nontriggered CT is extensively utilized. The study-purpose is to evaluate the in vitro agreement in coronary calcium score between nontriggered thoracic CT and ECG-triggered cardiac CT. METHODS Three artificial coronary arteries containing calcifications of different densities (high, medium, and low), and sizes (large, medium, and small), were studied in a moving cardiac phantom. Two 64-detector CT systems were used. The phantom moved at 0-90 mm∕s in nontriggered low-dose CT as index test, and at 0-30 mm∕s in ECG-triggered CT as reference. Differences in calcium scores between nontriggered and ECG-triggered CT were analyzed by t-test and 95% confidence interval. The sensitivity to detect calcification was calculated as the percentage of positive calcium scores. RESULTS Overall, calcium scores in nontriggered CT were not significantly different to those in ECG-triggered CT (p>0.05). Calcium scores in nontriggered CT were within the 95% confidence interval of calcium scores in ECG-triggered CT, except predominantly at higher velocities (≥50 mm∕s) for the high-density and large-size calcifications. The sensitivity for a nonzero calcium score was 100% for large calcifications, but 46%±11% for small calcifications in nontriggered CT. CONCLUSIONS When performing multiple measurements, good agreement in positive calcium scores is found between nontriggered thoracic and ECG-triggered cardiac CT. Agreement decreases with increasing coronary velocity. From this phantom study, it can be concluded that a high calcium score can be detected by nontriggered CT, and thus, that nontriggered CT likely can identify individuals at high risk of cardiovascular disease. On the other hand, a zero calcium score in nontriggered CT does not reliably exclude coronary calcification.