Low-dose spiral computed tomography of the thorax: comparison with the standard-dose technique.

RATIONALE AND OBJECTIVES This study was conducted to determine whether significant reduction in radiation dose (250 mAs-->50 mAs), in chest computed tomography (CT) using volume acquisition affects image quality or the detectability of pathologic findings in the lung and mediastinum. METHODS Phantom studies were conducted to evaluate resolution and noise level, then a patient study was then done. Fifty consecutive patients (10 normal and 40 abnormal) cases were examined. After a scan (250 mAs, 120OkVp) of the entire thorax, five additional slices (50 mAs, 120kVp) at the level of the abnormality were obtained. Three independent observers evaluated the visibility of normal lung and mediastinal structures, as well as image noise. The mean score was compared between the standard and low doses. In a second study, an independent evaluation of the presence or absence of pathologic findings was made by four observers. RESULTS Lucite phantom studies demonstrated diminished low-contrast resolution and increased noise level for the low-dose technique. Observers assessed more noise in the low-dose images (P < 0.001). The normal structures were judged to be more visible with standard dose (P < 0.01), but the magnitude of the judged differences was small especially in the lung. No differences were found in the detection of lung and mediastinal abnormalities (P > 0.10). CONCLUSIONS The low-dose technique was effective in demonstrating pathologic findings for the lung and mediastinum. Low-dose spiral CT should be considered as a viable alternative to standard-dose spiral CT.