Detection of blebs and bullae in patients with primary spontaneous pneumothorax by multi-detector CT reconstruction using different slice thicknesses.

INTRODUCTION

The aim of this study was to compare the diagnostic performances of multi-detector computed tomography (MDCT) reconstruction at two different slice thicknesses (1 mm, 'high resolution' vs. 5 mm, 'routine') with respect to the detection of blebs and bullae (BBs) in patients with primary spontaneous pneumothorax (PSP).

METHODS

Thirty-one patients underwent wedge resection of BBs (29 unilateral and 2 bilateral) for PSP from January 2010 to January 2013. Two observers assessed the presence and locations of BBs independently using high-resolution CT (HRCT) and routine CT reconstruction, and compared the sensitivities of each reconstruction method for BB detection using operative findings as a standard reference. In addition, the number of BBs in each CT image set was recorded and inter-observer agreements were evaluated.

RESULTS

Sensitivity for the detection of BBs was significantly better for HRCT than routine CT (97.0% vs. 63.6% for observer 1 and 94.0% vs. 57.6% for observer 2, respectively, both P-values < 0.001). On a per-bleb and a per-bulla basis, inter-observer agreements regarding BBs by HRCT were good and very good (k = 0.66 and 0.94, respectively) and superior to those determined by routine CT (k = 0.59 and 0.60, respectively).

CONCLUSION

Different slice thickness reconstructions influence the diagnostic efficacy of MDCT for the detection of BBs in patients with PSP. High-resolution thin slice CT reconstruction was found to have a significantly greater sensitivity than routine thicker slice thickness reconstruction for the detection of BBs.