Figure 1
Hounsfield units (HU) (iodine enhancement level) as lung iPBV values are calculated automatically in several patterns, including the whole lung (underlined).
Figure 2
Individual data of lung iPBV value with or without COPD: in early phase (a) and in late phase (b).
Table 1
Summary of results of analysis with and without COPD. Note: NS = not significant. FEV1/FVC = forced expiratory volume in 1 second/ forced vital capacity. %LAA-950 = low attenuation areas using a Hounsfield Unit (HU) threshold of −950.
| Patients with COPD(n = 14) | Patients without COPD(n = 48) | p value | |
|---|---|---|---|
| Age(years) | 70.4 ± 10.3 | 65.3 ± 15.3 | NS |
| Male,n | 8 | 13 | NS |
| Lung iPBV in early phase (HU) | 20.1 ± 5.5 | 30.6 ± 7.6 | < 0.001 |
| Lung iPBV in late phase (HU) | 12.3 ± 3.7 | 15.3 ± 4.6 | 0.051 |
| FEV1/FVC(%) | 63.1 ± 5.4 | 79.1 ± 7.1 | < 0.001 |
| %LAA-950(%) | 8.0 ± 3.8 | 4.0 ± 2.4 | < 0.001 |
Figure 3
ROC analyses demonstrated moderate discriminatory power for using lung iPBV to differentiate between patients with and without COPD. When less than 27.5 HU was used as the threshold for diagnosis, the sensitivity, specificity, positive predictive value, and negative predictive value were 92.3, 64.6, 41.4, and 96.8%, respectively.
Figure 4
The relationship between lung iPBV value and the percentage area of emphysema (%LAA-950) in all patients: in early phase (a) and in late phase (b). The lung iPBV values in early phase (R = −0.57, p = 0.0002) and late phase (R = −0.45, p = 0.005) had significant negative correlations with %LAA-950.