Fig 1.
Pulmonary nodules were simulated by injecting 0.3–0.8 mL of our agar-mixture into the lung parenchyma using a 5 mL syringe and a 20 G cannula.
Table 1.
Imaging parameters of the three MRI sequences.
Fig 2.
CT-MRI comparison of 2 nodules with a diameter of 2mm.
Fig 3.
Logistic regression analysis illustrating the estimated probabilities of nodule detection depending on the nodule’s diameter for each MR sequence and each reader.
The shaded areas around the lines indicate the 95% confidence interval.
Table 2.
MRI sensitivities (± standard deviation) for different nodule diameters according to logistic regression analysis.
Table 3.
MRI sensitivities by nodule size intervals.
Table 4.
False-positive findings for each sequence and reader.
Table 5.
Positive Predictive Value (PPV).
Fig 4.
Bland-Altman-plot showing the differences between the nodule size measurements using MRI vs. CT for each of the tested sequences and each reader.
The red lines indicate the bias (mean difference between CT and MR measurements), and the dashed blue lines indicate the 95% confidence interval (±1.96 × standard deviation).