Standardizing CT lung density measure across scanner manufacturers

Abstract

Purpose: Computed Tomography (CT) imaging of the lung, reported in Hounsfield Units (HU), can be parameterized as a quantitative image biomarker for the diagnosis and monitoring of lung density changes due to emphysema, a type of chronic obstructive pulmonary disease (COPD). CT lung density metrics are global measurements based on lung CT number histograms, and are typically a quantity specifying either the percentage of voxels with CT numbers below a threshold, or a single CT number below which a fixed relative lung volume, nth percentile, falls. To reduce variability in the density metrics specified by CT attenuation, the Quantitative Imaging Biomarkers Alliance (QIBA) Lung Density Committee has organized efforts to conduct phantom studies in a variety of scanner models to establish a baseline for assessing the variations in patient studies that can be attributed to scanner calibration and measurement uncertainty.

Methods: Data were obtained from a phantom study on CT scanners from four manufacturers with several protocols at various tube potential voltage (kVp) and exposure settings. Free from biological variation, these phantom studies provide an assessment of the accuracy and precision of the density metrics across platforms solely due to machine calibration and uncertainty of the reference materials. The phantom used in this study has three foam density references in the lung density region, which, after calibration against a suite of Standard Reference Materials (SRM) foams with certified physical density, establishes a HU-electron density relationship for each machine-protocol. We devised a 5-step calibration procedure combined with a simplified physical model that enabled the standardization of the CT numbers reported across a total of 22 scanner-protocol settings to a single energy (chosen at 80 keV). A standard deviation was calculated for overall CT numbers for each density, as well as by scanner and other variables, as a measure of the variability, before and after the standardization. In addition, a linear mixed-effects model was used to assess the heterogeneity across scanners, and the 95% confidence interval of the mean CT number was evaluated before and after the standardization.

Results: We show that after applying the standardization procedures to the phantom data, the instrumental reproducibility of the CT density measurement of the reference foams improved by more than 65%, as measured by the standard deviation of the overall mean CT number. Using the lung foam that did not participate in the calibration as a test case, a mixed effects model analysis shows that the 95% confidence intervals are [-862.0 HU, -851.3 HU] before standardization, and [-859.0 HU, -853.7 HU] after standardization to 80 keV. This is in general agreement with the expected CT number value at 80 keV of -855.9 HU with 95% CI of [-857.4 HU, -854.5 HU] based on the calibration and the uncertainty in the SRM certified density.

Conclusions: This study provides a quantitative assessment of the variations expected in CT lung density measures attributed to non-biological sources such as scanner calibration and scanner x-ray spectrum and filtration. By removing scanner-protocol dependence from the measured CT numbers, higher accuracy and reproducibility of quantitative CT measures were attainable. The standardization procedures developed in study may be explored for possible application in CT lung density clinical data.

Keywords: COPD; CT scanner calibration; Hounsfield Unit correction; Quantitative Imaging Biomarker; lung density CT; lung density SRM; lung density reference phantom.

Figure 1

CT slice image of the COPDGene 2 phantom used in this study. The CT numbers labeled are approximate from a typical 120 kV scan.

Figure 2

Results after executing procedures used to eliminate machine dependent variations for the 22 vendor‐protocols for the middle density foam (nominal density 192.6 kg/m³). Upper panel: trace plot for each of the 22 vendor‐protocols at each of the calibration steps. Lower panel: boxplot representation for each vendor‐protocol. The box represents the first and third quartile of the distribution, and the horizontal bar in the middle of the box is the median value. The upper and lower whiskers are the maximum and minimum values excluding outliers. The data points outside of the whiskers are considered outliers. The reference value (solid line) at 80 keV is calculated based on the SRM foam density and assumed composition, and the mean value (dashed line) is the unweighted mean of the HU value at 80 keV for the entire group. The maximum variation and standard deviation in HU at each step are also listed. All the data analysis and graphics were performed with R (http://www.R-project.org).

Figure 3

Boxplot showing distribution of the HU value after scanner recalibration of all kVp and mGy settings, mapped to a single energy of 80 keV, plotted by vendor. Foams 1 to 3 are the reference foams, and the lung foam is the foam backing that fills the entire phantom. The lung foam did not participate in the calibration fit, and therefore serves as a test case to assess the variation.

Figure 4

CT number variation by dose levels for each foam, labeled as Foam 1, 2, 3, and Lung (L), for calibrated CT number at 80 keV. The number of data points for each of the dose levels (1.5 mGy, 3 mGy, and 6 mGy) is 5, 9, and 8 respectively. The expected reduction in the SD due to increased dose rate is observed for foams 1 and 3. With the exception of the lung foam at 3 mGy, all the SD's are below 1 HU. The high value for the lung foam at 3 mGy is mostly due to a single scanner, as indicated in the 4th panel of Fig. 3.

Figure 5

Histograms showing the relative distributions of the HU values for all scanner‐protocols as measured (light) and corrected to 80 keV (dark) for each of the 4 densities. The more unimodal distribution for all densities illustrates the level of standardization achieved.

Figure 6

Forest Plots: Left panel – CT number as measured (scaled H _raw ); right panel – CT number after the recalibration standardization procedures (scaled H _80keV ), for the backing lung foam in the phantom (which served as an “unknown”), showing the reduction in variability and improved accuracy. The solid black dots and horizontal thin black line segments with an “x” symbol on both ends represent the measured values and their associated expanded uncertainties (95% confidence). The open squares and horizontal thick gray line segments with a “+” symbol on both ends represent the corresponding values predicted by the fitted models and their associated expanded uncertainties (95% confidence). The large solid diamond at the bottom and the vertical, thin solid line indicate the overall mean (consensus value) determined by a mixed effects model, in both cases fitted to the data taking the effects of scanner, kVp and mGy into account. In this mixed effects model, the overall mean and dose (mGy) are the fixed effects, and scanner and kVp (which is nested within scanner) are the random effects. The expected CT number value at 80 keV is −855.9 HU with 95% CI of [−857.4 HU, −854.5 HU] (based on the propagation of uncertainty in the SRM certified density). This is in general agreement with the information conveyed by the 95% confidence intervals of [−862.0 HU, −851.3 HU] (before standardization) and [−859.0 HU, −853.7 HU] (after standardization), shown as the error bars for the overall mean at the bottom.

Figure 7

Expected change in CT number for different kinds of tissue due to a 1% change in the calibration parameter α, calculated based on the ICRP (1972) tissue compositions, to illustrate how the calibration error affects the CT number variation, the degree of whcih depending on the electron density and effective atomic number of the particular tissue.

Figure 8

Stoichiometric H/C ratio of various possible polyurethane compositions. Dash lines are the upper and lower bound of the 95% confidence interval of the mean value as determined by PGAA measurement. Those compositions that fall within this interval are considered as possibilities, and are used to assess the error associated with the composition.