Efficacy of Computer-Aided Detection System and Thin-Slab Maximum Intensity Projection Technique in the Detection of Pulmonary Nodules in Patients With Resected Metastases

Abstract

To evaluate the efficacy of the computer-aided detection (CAD) system and thin-slab maximum intensity projection (MIP) technique in the detection of pulmonary nodules at multidetector computed tomography (CT) in patients who underwent metastatectomy. This retrospective study was approved by the institutional review board and patients' informed consent was waived. Forty-nine consecutive patients who underwent pulmonary metastatectomy were enrolled. Four chest radiologists analyzed preoperative 1-mm section CT images and recorded the locus of each nodule candidate. Afterward, they reevaluated the images once using CAD software and once with thin-slab MIP given the results of 1-mm section CT alone. The reference standard for nodule presence was established by a consensus panel and pathologic records for malignant nodules. A total of 514 nodules were identified by a consensus panel. Of 212 nodules surgically removed, 121 nodules were malignant. The sensitivity of each observer in detecting malignant nodules with thin-section CT scans alone was 91%, 88%, 87%, and 86% for observers A- to D, respectively. With CAD, sensitivity increased significantly to 95%, 95%, 94%, and 95% (P< 0.05 for observer B-D), and using MIP increased to 94%, 96%, 91%, and 92% (P < 0.05 for observer B-D), respectively. There were no significant differences in sensitivity between CAD and MIP for the detection of malignant nodules. The average number of false-positive findings per patient was 0.8 with thin-section CT alone, 1.1 with CAD, and 1.4 with MIP. In candidates for metastatectomy, reading with the aid of either CAD or MIP significantly improved the detection of malignant nodules compared with using thin-section CT alone.