Study objective: Computed tomography provides measurements of lung attenuation which reflect changes in the air to tissue ratio and can thereby be employed for diagnosis of diffuse lung disease. In this prospective study, we quantitatively analyzed lung density by high resolution computed tomography (HRCT) in 26 healthy volunteers, 15 patients with chronic obstructive pulmonary disease (COPD), and 15 patients with idiopathic lung fibrosis (IPF). The procedure was standardized by examination of 3 scans at the carina +/- 5 cm and by defining inflation levels by %VC using an on-line hand held spirometer.
Results: Performance of HRCT at 50% VC provides not only significant and distinguishable group data, but is the easiest to carry out for dyspneic patients. The mean lung density at 50% VC for healthy subjects was -820 +/- 4.2 (mean +/- SEM) Hounsfield units (HU). It was significantly lower (p < 0.01) in COPD patients (-865 +/- 9.2 HU), and considerably higher (-697 +/- 17.8 HU, p < 0.001) in the IPF group. At an inflation level of 20% VC, mean lung density values were similarly distributed, at significantly lower values relative to those at 50% VC, but the procedure was more difficult to perform for patients with dyspnea. In contrast, at 80% VC, lung density values for the COPD and control groups were not significantly different (p = 0.08). The sensitivity to detect COPD was improved by selecting HRCT values lower than -900 HU, which represent the part of the lung with an increased air/tissue ratio. For IPF patients an increase of lung density values above -699 HU was characteristic, indicating a decrease of the air/tissue relationship.
Conclusion: From our data we propose to perform quantitative HRCT measurements at 50% VC. Diagnosis of diffuse lung disease can be further improved by consideration of specific CT -value intervals. Spirometrically controlled quantitative HRCT is a clinically meaningful tool for the assessment of diffuse parenchymal lung disease.