As shown in Figure 3, both normal (standard) and sharp reconstruction algorithms yielded nearly identical measurements of lung density for the same image (slope = 1.006, r = 0.99). Thus we used the sharp (high resolution) reconstruction algorithm for both airway CSA and lung density measurements.
An example of the CT images obtained from the same scan level at the four time points (before metha-choline injection, and 30 s, 2 min, and 4 min after methacholine injection) is depicted in Figure 4. The insert to the upper right of each image demonstrates the constriction of a single airway. Maximal broncho-constriction was always found at the first time point evaluated, at approximately 30 s after methacholine bolus, with the airway CSA decreasing by 31 ± 3% (mean ± SEM) from baseline (p < 0.001; Fig 5). The CSAs at 2 min and 4 min were not significantly different from baseline values, although they tended to remain lower. No significant statistical difference was observed between the different groups of airways in any of the time points. Pulmonary Infection
The density measurements (percentage of air content) showed no difference between the peripheral and central lung regions. Thus, all regions, both decreased air content. The same pattern was observed in all other experiments. Figure 9 summarizes the positional gradients of lung density for all experiments, showing the relationship between lung density as expressed by percentage of air content vs (log) lung height at various time points. At 30 s, during maximal bronchoconstriction, there is, in general, an increase in air content that is greater at the dependent lung regions. At 2 min and 4 min, the air content returned to baseline levels, paralleling the time changes that we observed with the airways CSA. The various parameters that summarize the density measurements and indexes of the relationship between lung density and lung height for all the experiments are presented in Table 1.
The results of the present study demonstrated that within the time frame used in this investigation, the central and peripheral, were used for further analysis. A logarithmic relationship between the lung percentage of air content and the lung height was demonstrated in all pigs, at each time point. Figure 6 illustrates this relationship in one of the pigs at baseline. Air content in this pig increases with the transition from dependent to nondependent regions (Fig 6, top, A). Using a logarithmic scale for the lung height (Fig 6, bottom, B), the slope of this relationship was 38.35% air per centimeter of lung height, and the Y-intercept (percentage of air) was 5.32. Figure 7 shows this relationship in the same pig during all experimental time points. At 30 s, the relationship substantially changed from baseline, particularly in the dependent lung regions (Fig 7, top, A). On a logarithmic scale (Fig 7, bottom, B), the slope was now 28.52% air per centimeter of lung height and the intercept was 24.46% air. Of interest is the fact that the variability of the data also significantly increased at 30 s. At 2 min and 4 min, the air content measurements returned to baseline levels. Figure 8 plots these data by relating the percentage of air content of the same regions (ROIs) of this pig at various time points after methacholine bolus to their baseline percentage of air content values. This plot clearly demonstrates that in most of the regions, air content increased at 30 s (ie, during maximal bronchoconstriction), thus deviating from the line of identity. Again note that the variability of percentage of air content measurements at 30 s was significantly greater than that observed at other time points (p < 0.001). The figure also illustrates that while in general there was increase in air content in 30 s compared to baseline, a few regions showed lung density measurements may be used as an index of peripheral changes occurring in response to bron-choconstrictive agents correlating to measurable structural responses. These measurements allow for assessment of lung structure and function beyond the spatial resolution of CT scanning techniques.
Figure 3. Correlation between normal and sharp algorithms for lung density measurements.
Figure 4. Images obtained at the four time points (before methacholine [baseline], and 30 s, 2 min, and 4 min after methacholine injection) in one of the experiments. The insert to the upper right of each image demonstrates the changes in CSA of a single airway.
Figure 5. Mean ( ± SEM) temporal changes in airway CSA in all experiments (n = 7). No significant statistical difference was observed between the different groups of airways in any of the time points. Changes were significant (p < 0.001) at 30 s vs all other time points. See Figure 1 for expansion of abbreviation.
Figure 6. Relationship between percentage of air content and the lung height in one of the experiments at baseline (top, A). The best fitted curve for this relationship is also plotted. The same relationship can be depicted as a linear function when using a logarithmic scale for the lung height axis (bottom, B).
Figure 7. Relationships between percentage of air content and the lung height in one of the experiments at baseline, 30 s (dashed line), 2 min, and 4 min after methacholine injection (top, A: linear scale for the lung height axis; bottom, B: logarithmic scale).
Figure 8. Percentage of air content of the same regions at various time points after methacholine injection in relation to their baseline values in one of the experiments. See Figure 1 for expansion of abbreviation.
Figure 9. Relationships between lung density (% air content) and lung height at various time points for all the experiments.
Table 1—Lung Density (Percentage of Air Content) Indexes and the Parameters of the Relationship Between Lung Density and Lung Height
|Variables||Baseline||30 s||2 min||4 min|
|Air content, %||65.87 ± 1.40||70.21 ± 1.65t||65.24 ± 1.54||66.50 ± 1.50|
|CV, %||8.60 ± 0.60||10.57 ± 0.68t||9.38 ± 0.70||9.14 ± 0.69|
|Slope||38.57 ± 2.96||28.29 ± 3.57t||38.48 ± 2.07||36.88 ± 2.99|
|Intercept, % air content||7.26 ± 5.80||29.45 ± 6.78t||5.56 ± 4.79||9.61 ± 6.58|
|Correlation coefficient, r||0.91 ± 0.01||0.74 ± 0.05t||0.87 ± 0.01||0.90 ± 0.01|