Objective
Dual-energy X-ray computed tomography (DECT) offers visualization of the airways and quantitation of regional pulmonary ventilation using a single breath of inhaled xenon gas. In this study we seek to optimize scanning protocols for DECT xenon gas ventilation imaging of the airways and lung parenchyma and to characterize the quantitative nature of the developed protocols through a series of test-object and animal studies.
Materials and Methods
The Institutional Animal Care and Use Committee approved all animal studies reported here. A range of xenon-oxygen gas mixtures (0, 20, 25, 33, 50, 66, 100%; balance oxygen) were scanned in syringes and balloon test-objects to optimize the delivered gas mixture for assessment of regional ventilation while allowing for the development of improved three-material decomposition calibration parameters. Additionally, to alleviate gravitational effects on xenon gas distribution, we replaced a portion of the oxygen in the xenon/oxygen gas mixture with helium and compared gas distributions in a rapid-prototyped human central-airway test-object. Additional syringe tests were performed to determine if the introduction of helium had any effect on xenon quantitation. Xenon gas mixtures were delivered to anesthetized swine in order to assess airway and lung parenchymal opacification while evaluating various DECT scan acquisition settings.
Results
Attenuation curves for xenon were obtained from the syringe test objects and were used to develop improved three-material decomposition parameters (HU enhancement per percent xenon: Within the chest phantom: 2.25 at 80kVp, 1.7 at 100 kVp, and 0.76 at 140 kVp with tin filtration; In open air: 2.5 at 80kVp, 1.95 at 100 kVp, and 0.81 at 140 kVp with tin filtration). The addition of helium improved the distribution of xenon gas to the gravitationally non-dependent portion of the airway tree test-object, while not affecting quantitation of xenon in the three-material decomposition DECT. 40%Xe/40%He/20%O2 provided good signal-to-noise, greater than the Rose Criterion (SNR > 5), while avoiding gravitational effects of similar concentrations of xenon in a 60%O2 mixture. 80/140-kVp (tin-filtered) provided improved SNR compared with 100/140-kVp in a swine with an equivalent thoracic transverse density to a human subject with body mass index of 33. Airways were brighter in the 80/140 kVp scan (80/140Sn, 31.6%; 100/140Sn, 25.1%) with considerably lower noise (80/140Sn, CV of 0.140; 100/140Sn, CV of 0.216).
Conclusion
In order to provide a truly quantitative measure of regional lung function with xenon-DECT, the basic protocols and parameter calibrations needed to be better understood and quantified. It is critically important to understand the fundamentals of new techniques in order to allow for proper implementation and interpretation of their results prior to wide spread usage. With the use of an in house derived xenon calibration curve for three-material decomposition rather than the scanner supplied calibration and a xenon/helium/oxyge...