Table 3. Sensitivity, Specificity, and Accuracy for Detection of SPLs on MDCT and MRI.
Reviewer 1 | P* | Reviewer 2 | P* | |||
---|---|---|---|---|---|---|
CT | MRI | CT | MRI | |||
Detection of SPL per se | ||||||
Sensitivity | 92.7% (179/193) | 97.9% (189/193) | 0.031 | 90.7% (175/193) | 99.5% (192/193) | < 0.001 |
PDAC | 92.1% (117/127) | 99.2% (126/127) | 0.012 | 89.0% (113/127) | 99.2% (126/127) | < 0.001 |
NET | 95.3% (41/43) | 93.0% (40/43) | 1.000 | 95.3% (41/43) | 100.0% (43/43) | 0.500 |
SPT | 100.0% (10/10) | 100.0% (10/10) | - | 100.0% (10/10) | 100.0% (10/10) | - |
Mass-forming AIP | 71.4% (5/7) | 100.0% (7/7) | 0.500 | 71.4% (5/7) | 100.0% (7/7) | 0.500 |
Metastasis | 100.0% (6/6) | 100.0% (6/6) | - | 100.0% (6/6) | 100.0% (6/6) | - |
Specificity | 100.0% (52/52) | 100.0% (52/52) | - | 100.0% (52/52) | 100.0% (52/52) | - |
Accuracy | 94.3% (231/245) | 98.4% (241/245) | 0.031 | 92.7% (227/245) | 99.6% (244/245) | < 0.001 |
Detection of SPL in consideration of secondary features | ||||||
Sensitivity | 97.9% (189/193) | 97.9% (189/193) | 1.000 | 97.4% (188/193) | 99.5% (192/193) | 0.219 |
PDAC | 99.2% (126/127) | 99.2% (126/127) | 1.000 | 98.4% (125/127) | 99.2% (126/127) | 1.000 |
NET | 95.3% (41/43) | 93.0% (40/43) | 0.500 | 95.3% (41/43) | 100.0% (43/43) | 0.500 |
SPT | 100.0% (10/10) | 100.0% (10/10) | - | 100.0% (10/10) | 100.0% (10/10) | - |
Mass-forming AIP | 85.7% (6/7) | 100.0% (7/7) | 1.000 | 85.7% (6/7) | 100.0% (7/7) | 1.000 |
Metastasis | 100.0% (6/6) | 100.0% (6/6) | - | 100.0% (6/6) | 100.0% (6/6) | - |
Specificity | 96.2% (50/52) | 98.1% (51/52) | 1.000 | 100.0% (52/52) | 100.0% (52/52) | - |
Accuracy | 97.6% (239/245) | 98.0% (240/245) | 1.000 | 98.0% (240/245) | 99.6% (244/245) | 0.219 |
*p-value was obtained between CT and MRI using McNemar test. AIP = autoimmune pancreatitis, MDCT = multidetector computed tomography, NET = neuroendocrine tumor, PDAC = pancreatic ductal adenocarcinoma, SPL = solid pancreatic lesion, SPT = solid pseudopapillary tumor