Detection of nodular lesions in the lung helical computed tomography: comparison between 360 degrees and 180 degrees reconstruction algorithms.

ABSTRACT

The present study has evaluated helical computed tomography (HS-CT) performed at a fast couch-top sliding speed (20 mm/sec, sufficient to cover the entire lung field during the holding of a single breath) with regard to its ability to detect pulmonary nodules in patients with metastatic lung disease. The detectability of lesions by HS-CT was compared between 360 degrees and 180 degrees reconstruction algorithms. The subjects were eight consecutive patients with metastatic lung tumors, in whom 214 lesions had been detected by conventional CT. The average diameter of the nodules detected was 7.7 mm (2-22 mm). To scan the entire lung, we first used conventional CT (10-mm slice thickness, 10-mm stepping) and then HS-CT with a couch-top speed of 20 mm/sec and a tube current of 50 mA. Both 360 degrees and 180 degrees algorithms were used to reconstruct images at intervals of 20 and 10 mm from HS-CT data. Detection rates of 79% for 20-mm images and 94% for 10-mm images were achieved with the 360 degrees algorithm. Detection rates of 81% for 20-mm images and 99% for 10-mm images were achieved with the 180 degrees algorithm. The detection of nodules was greater for 10-mm images than 20-mm images (P < 0.01). There was no significant difference between the 360 degrees and 180 degrees algorithms. Images acquired using the 180 degrees algorithm, however, depicted lesions with greater sharpness than those acquired using the 360 degrees algorithm. It is presumed that they indirectly show the superiority of the 180 degrees algorithm over the 360 degrees algorithm. HS-CT should prove a useful screening technique for lesions in the lung field.