Computer-aided of lung nodules on thin collimation MDCT: impact on radiologists’ performance

etection, specific diagnosis, and effective treatment of a large number of lung diseases, including primary or secondary bronchopulmonary cancers, begin by identifying lung nodules. CT sensitivity in detecting lung nodules is much better than chest x-ray sensitivity. In addition, the helical scanner is significantly better than conventional CT (1). Today, with the increasingly widespread use of multidetector CT scanners, the lungs can be acquired in their entirety during a single breath hold with high spatial resolution resulting from thin collimation and overlapping slices in reconstruction. High-resolution acquisition provides anatomic precision of the bronchovascular tree, which results in finer detection of small lung nodules. This detailed anatomy is obtained, however, at the cost of a large number of images to interpret, which makes reading long and laborious, with an increased risk of errors in detecting nodules. Therefore, despite the technological progress made in CT imaging, a high number of lung nodules go unrecognized by radiologists (2-4). To prevent detection errors, several computer-aided Résumé Abstract