Spiral CT evaluation of pulmonary emphysema using a low-dose technique

To evaluate the diagnostic accuracy and clinical acceptability of low-dose spiral CT for determining pulmonary volumes and emphysema extension in patients with pulmonary emphysema, in comparison with studies based on spiral CT at conventional dose.

We prospectively evaluated eighteen patients, current or former smokers, with a clinical diagnosis of chronic obstructive pulmonary disease. All the patients underwent: HRCT with three scans at predetermined levels; quantitative spiral CT, with two inspiratory scans, one conventional scan at 240 mA, and the second one a low-dose scan at 80 mA. We used the following parameters: 120 kV, rotation time 0.8", scan time less than 20" (single inspiratory breath-hold), layer thickness 7.5 mm, pitch 6 (high speed), interpolation algorithm at 180 degrees. A 3D reconstruction was performed, with segmentation of the lungs and automatic quantification of pulmonary volumes. We compared the volumes of absolute and percent emphysema and the ratings of the dose delivered to the patient (CTDIw and DLP) obtained with the two spiral CT scans with each other and with the respiratory function tests.

The average total lung capacity (TLC) obtained by conventional-dose spiral CT (CTs1) was 6889.4 cc (SD +/-1813.2), and the capacity with low-dose spiral CT (CTs2) was 6929.4 cc (SD +/-1811.6). The percentage of emphysema was 39.7% (range: 2.2-63.5%; SD: +/-19.9) for the CTs1 and 41.1% (range: 2.1-66.4%; SD: +/-20). The CTDIw corresponding to CTs1 was 12.2 mGy (range: 11.9-16.4; SD: +/-1), the one corresponding to CTs2, 3.6 mGy (range: 3.6-4.9; SD: +/-0.3). The DLP corresponding to CTs1 was 391.7 mGy x cm (range: 333.3-518.9; SD: +/-46.7), the one corresponding to CTs2 was 117.8 mGy x cm (range: 100.3-156; SD: +/-14). As for the respiratory function tests, the total lung capacity (TLC) obtained by body plethysmography was 7061 cc (SD: +/-2029.7); the percent TLC was 115.9 (range: 66-165; SD: +/-27.6), the forced expiratory volume at one second (FEV1%, percentage of predicted value) was 46.7% (range: 17-123; SD: +/-27.3), residual volume (RV%) as a percentage of predicted value was 186.3 (range: 84-359; SD: +/-80.7), the Tiffeneau index (TI) was 46% (range: 25-71; SD: +/-15.7). We observed a very significant correlation between radiological and functional TLC for both CT methods. The percentage scores for emphysema obtained with the two methods correlated significantly with the functional indexes. The pixel index of CTs1 correlated with TLC% (r=0.87; p<0.0001), FEV1% (r=-0.53; p<0.02), RV% (r=0.76; p=0.004), TI (r=-0.79; p=0.0001). The pixel index of CTs2 correlated with TLC% (r=0.87; p<0.0001), FEV1% (r=-0.56; p=0.01), RV% (r=0.78; p=0.003), TI (r=-0.8; p=0.0001). The adoption of the method with low tube current entailed a highly significant reduction in the estimated dose delivered to patients (CTDIw and DLP) with r=0.9 and p < 0.0001.

Quantitative low-dose spiral CT is a very good method to quantify pulmonary volumes and calculate the extension of the anatomic emphysema. The reduction of mA from 240 to 80 lowers the estimated dose by 30%, without compromising the accuracy of the results. Our study achieved a highly significant correlation between the results obtained with the two spiral CT techniques and between these results and the respiratory function tests. In clinical practice, the easiest way to reduce the dose in spiral CT of the lung is to reduce the tube current. The low-dose method allows a significant reduction in radiation exposure. Further studies are required to establish to what extent the dose can be reduced without increasing in quantum noise and thereby compromising the quality of the study.