Computer-assisted detection of pulmonary nodules: evaluation of diagnostic performance using an expert knowledge-based detection system with variable reconstruction slice thickness settings

Abstract

The purpose of this study was to evaluate the performance of a computer-assisted diagnostic (CAD) tool using various reconstruction slice thicknesses (RST). Image data of 20 patients undergoing multislice CT for pulmonary metastasis were reconstructed at 4.0, 2.0 and 0.75 mm RST and assessed by two blinded radiologists (R1 and R2) and CAD. Data were compared against an independent reference standard. Nodule subgroups (diameter >10, 4-10, <4 mm) were assessed separately. Statistical methods were the ROC analysis and Mann-Whitney U test. CAD was outperformed by readers at 4.0 mm (Az = 0.18, 0.62 and 0.69 for CAD, R1 and R2, respectively; P<0.05), comparable at 2.0 mm (Az = 0.57, 0.70 and 0.69 for CAD, R1 and R2, respectively), and superior using 0.75 mm RST (Az = 0.80, 0.70 and 0.70 and sensitivity = 0.74, 0.53 and 0.53 for CAD, R1 and R2, respectively; P<0.05). Reader performances were significantly enhanced by CAD (Az = 0.93 and 0.95 for R1 + CAD and R2 + CAD, respectively, P<0.05). The CAD advantage was best for nodules <10 mm (detection rates = 93.3, 89.9, 47.9 and 47.9% for R1 + CAD, R2 + CAD, R1 and R2, respectively). CAD using 0.75 mm RST outperformed radiologists in nodules below 10 mm in diameter and should be used to replace a second radiologist. CAD is not recommended for 4.0 mm RST.

Computer-assisted detection of pulmonary embolism: performance evaluation in consensus with experienced and inexperienced chest radiologists

“…Similar to other clinical settings such as the detection of small pulmonary nodules, in which CAD has been used successfully, CAD of intra-arterial emboli may assist radiologists in PE detection by reducing the chance of missing PE patients [40][41][42][43][44][45][46][47]. It is notable, that in automated detection of lung nodules CAD systems are now established "second reader" diagnostic tools, significantly enhancing the performance of radiologists in consensus [47,48].…”

Pulmonary nodule detection on MDCT images: evaluation of diagnostic performance using thin axial images, maximum intensity projections, and computer-assisted detection

“…Most of the studies using CAD were performed on CT scans with different slice thickness and reconstruction increments. In studies evaluating CAD systems, slice thickness varied from 10 mm to 0.75 mm [19][20][21][22][23][24][25][26][27][28][29][30][31][32][33]. The minimal size of nodules depends on the studies' design and on the parameter of the acquisition.…”

“…Therefore, several approaches have been proposed to improve lung nodule detection: (1) monitor viewing using cine-mode [8,9]; (2) double independent reading [10,11]; (3) maximum intensity projections (MIP) [12][13][14][15][16]; (4) computer-assisted detection (CAD), which can guide the radiologist to questionable structures. Previous investigations using CAD with CT have showed an increase of pulmonary nodule detection sensitivity by radiologists [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35].…”

Value of axial and coronal maximum intensity projection (MIP) images in the detection of pulmonary nodules by multislice spiral CT: comparison with axial 1-mm and 5-mm slices

“…To date, a number of methods and display modes such as cine-based viewing [10], multiplanar reformations [11], computer aided diagnosis (CAD) [12][13][14][15], and the MIP technique [5][6][7][8] have been used and evaluated in order to enhance the radiologists' performance in pulmonary nodule detection and approach the problem of prohibitive image numbers. The effectiveness of double reading to improve sensitivity [16] and the influence of reader experience [15] have also been subject to recent research.…”