Table 2:
Metrics of rCBV estimation methodsa
| Method | RC (RCL–RCU) | wCV | tSD | bSD | wSD |
|---|---|---|---|---|---|
| A) Normalized rCBV (tumor) | |||||
| 2 (CTI) | 1.78 (1.44–2.34) | 0.31 | 0.96 | 0.71 | 0.64 |
| 3 (GV) | 1.95 (1.57–2.57) | 0.25 | 1.04 | 0.77 | 0.70 |
| 6 (NEI) | 2.07 (1.67–2.73) | 0.24 | 1.57 | 1.38 | 0.75 |
| 1 (UTI) | 3.16 (2.55–4.16) | 0.39 | 1.55 | 1.05 | 1.14 |
| 5 (MSD) | 3.54 (2.86–4.66) | 0.56 | 1.30 | 0.25 | 1.28 |
| 4 (PBC) | 3.93 (3.17–5.18) | 0.67 | 1.85 | 1.18 | 1.42 |
| B) Standardized rCBV (tumor) | |||||
| 2 (CTI) | 3869 (3121–5093) | 0.18 | 3142 | 2814 | 1397 |
| 4 (PBC) | 4020 (3243–5292) | 0.20 | 3045 | 2677 | 1451 |
| 3 (GV) | 4214 (3399–5547) | 0.17 | 3163 | 2773 | 1521 |
| 1 (UTI) | 4261 (3437–5609) | 0.19 | 2958 | 2526 | 1538 |
| 5 (MSD) | 4591 (3703–6043) | 0.13 | 6437 | 6220 | 1658 |
| 6 (NEI) | 5250 (4235–6911) | 0.16 | 5732 | 5410 | 1895 |
Note:—ΔR2*(t) indicates methods for which the transverse relaxation time courses are computed from the signal time courses; UTI, integration of ΔR2*(t) with the trapezoid rule; CTI, integration of ΔR2*(t) with the trapezoid rule and correction for T1 and T2 extravasation effects; GV, γ-variate fit to ΔR2*(t); NEI, negative enhancement integral of S(t); MSD, maximum signal drop of S(t); PBC, postbolus baseline (recirculation) correction; RCL, lower RC confidence interval; RCU, upper RC confidence interval; tSD, total SD; bSD, between-subject SD; wSD, within-subject SD.
Shown in A and B are the RC, upper and lower 95% CI for RC, wCV, and SDs (total, between, and within-subject) for nRCBV and sRCBV, respectively, in which methods are sorted in order of greatest repeatability, as determined by the RC for nRCBV or sRCBV.