TABLE VI.
U-test comparison of ratings between real and simulated tumors for each observer, testing for the hypothesis that real and simulated tumors have the same score distributions. AUC values are shown for each individual reader and for a MRMC analysis.
| U-test p values | Mean AUC (95% CI) | ||||
|---|---|---|---|---|---|
| Reader | Years of experience | Nonspiculated | Spiculated | Nonspiculated | Spiculated |
| Reader 1 | 2 | 0.3 | 0.861 | 0.57 (0.42, 0.69) | 0.49 (0.38, 0.60) |
| Reader 2 | 8 | 0.752 | 0.006 | 0.52 (0.37, 0.67) | 0.69 (0.58, 0.78) |
| Reader 3 | 4 | 0.670 | 0.164 | 0.47 (0.32, 0.62) | 0.57 (0.47, 0.68) |
| Reader 4 | 4 | 0.709 | 0.049 | 0.47 (0.34, 0.63) | 0.61 (0.50, 0.71) |
| Reader 5 | 39 | 0.009 | 0.136 | 0.68 (0.55, 0.78) | 0.58 (0.47, 0.68) |
| MRMC | — | — | — | 0.54 (0.44, 0.65) | 0.58 (0.50, 0.68) |