Reducing Radiation Dose in Quantitative CT of Obstructive
Lung Disease By Sean B. Fain, PhD
Chronic obstructive pulmonary disease (COPD) and asthma are the most common chronic lung diseases, affecting an estimated 38 million people in the U.S. Noninvasive imaging of lung disease using X-ray CT has long been the clinical standard for diagnosing lung diseases and their regional expression. Clinical lung CT uses a standard scale of Hounsfield units (HU) making it feasible to set thresholds to identify low density lung parenchyma. Typically alveolar tissue loss in emphysema is best seen during full inspiration with regional patterns reflecting COPD subtypes. Similarly, small airway disease in asthma leads to low density as a result of hyperinflation and air trapping. In both diseases, increased central airway thickening due to bronchiectasis and remodeling are also important indicators of disease subtypes. 
Several quantitative CT (qCT) measures have been introduced in the radiology and pulmonary medicine literature to numerically characterize phenotypic expression and severity of both COPD and asthma. The fractional extent of air in the lung volume below a HU threshold, typically -950 or -910 HU, has been shown to correlate with degree of emphysema on histology . More recently, measures of airway wall thickness have been shown to correlate with histology from bronchoscopic biopsy in asthma . However, several confounding factors can influence both HU and airway dimension measurements including varying lung inflation volume, image noise from poor photon statistics, and differing reconstruction and scatter correction methods used by CT vendors. The QIBA COPD-Asthma Technical Committee is charged with understanding the sources of bias and variance in these measurements and developing methods that are comparable across imaging systems and system configurations.
These tasks are complicated by increased public sensitivity to X-ray radiation dose from serial CT scans for research subjects with early lung disease. The challenge of achieving high spatial resolution while maintaining accurate measures of quantitative density argues for a new approach, in part because low-density HU values in the lung parenchyma are sensitive to noise, scatter correction, and helical artifact. Typically, qCT protocols have limited these variations by carefully controlling selection of acquisition protocol and reconstruction parameters, slowing incorporation of dose-saving algorithms such as automatic exposure control (AEC) and iterative image reconstruction (IR).
The purpose of a recently funded QIBA contract tested whether AEC and IR can reduce radiation dose while maintaining quantitative measures. Our approach uses the known reference standards in the COPDGene® phantom  developed in collaboration with the National Institute for Standards and Technology (NIST) [Fig. 1] on a 64-slice dual-energy CT scanner (GE 750 HD). This enables comparison to known values under controlled conditions. Among the key results are the consistency of HU values with an approximately 25% reduction in dose using AEC [Table 1], and the improved quantitative accuracy of wall thickness measures with adaptive statistical IR (ASIR). In the latter, ASIR was performed with a high frequency reconstruction kernel and increased spatial sampling—achieved by a reduced display field of view (DFOV). A combination of AEC with IR reconstruction using higher frequency kernels and reduced DFOV shows promise. We are currently testing this revised reconstruction protocol with raw projection data derived from human subjects and across multiple vendor platforms. The QIBA COPD/Asthma Technical Committee is working towards a protocol to accurately measure CT number for lung density and airway morphology with a single low-dose inspiratory helical CT scan.
Figure 1. COPDGene®2 test object with cylindrical foam density standards and acrylic tubes mimicking airway wall and lumen. Densities are indicted below in Table 1.
 Thurlbeck WM, Muller NL. Emphysema: Definition, Imaging, and Quantification. American Journal of Roentgenology. 1994;163(5):1017-25. Epub 1994/11/01. doi: 10.2214/ajr.163.5.7976869. PubMed PMID: 7976869.
 Hayhurst MD, MacNee W, Flenley DC, Wright D, McLean A, Lamb D, et al. Diagnosis of Pulmonary Emphysema by Computerised Tomography. Lancet. 1984;2(8398):320-2. Epub 1984/08/11. PubMed PMID: 6146866.
 Aysola RS, Hoffman EA, Gierada D, Wenzel S, Cook-Granroth J, Tarsi J, et al. Airway Remodeling Measured by Multidetector CT is Increased in Severe Asthma and Correlates with Pathology. Chest. 2008;134(6):1183-91. Epub 2008/07/22. doi: 10.1378/chest.07-2779. PubMed PMID: 18641116; PubMed Central PMCID: PMC2859729.
 FDA Challenges and Opportunities Report,March 2004. Innovation or Stagnation: Challenge and Opportunity on the Critical Path to New Medical Products. [online reference]
 Sieren JP, Newell JD, Judy PF, Lynch DA, Chan KS, Guo J, et al. Reference Standard and Statistical Model for Intersite and Temporal Comparisons of CT Attenuation in a Multicenter Quantitative Lung Study. Medical Physics. 2012;39(9):5757-67. Epub 2012/09/11. doi: 10.1118/1.4747342. PubMed PMID: 22957640; PubMed Central PMCID: PMC3448623.
Sean B. Fain, PhD, is an associate professor in the Medical Physics Department at the University of Wisconsin-Madison and a member of the QIBA COPD-Asthma Technical Committee. As director of the Image Analysis Core Facility, he leads MR imaging and CT studies for quantitative assessment of asthma severity and progression.
[ BACK TO TOP ]