Table 2.
Summary of studies comparing quantitative myocardial perfusion with MRI and PET
| Study | Study population | Sequential/simultaneous PET and MRI scans | PET tracer | MRI perfusion acquisition approach | MRI perfusion analysis model |
|---|---|---|---|---|---|
| Pärkkä et al.22 | 18 healthy males |
Sequential Separate days |
15O-water | Single bolus, single sequence | 1TCM |
| Fritz-Hansen et al.20 | 10 healthy males |
Sequential 3 - 12 days between MRI and PET |
13N-ammonia | Single bolus, single sequence | 1TCM |
| Pack et al.37 | 4 healthy volunteers, one heart-transplant patient |
Sequential 1–6 months between MRI and PET |
13N-ammonia | Single bolus, single sequence | Independent deconvolution |
| Morton et al.21 | 38 patients with known or suspected CAD |
Sequential 3 ± 6 days between PET and MRI |
13N-ammonia | Dual bolus | Fermi deconvolution |
| Qayyum et al.23 | 14 patients with CAD |
Sequential 6.6 ± 30.3 days |
82Rb | Single bolus, single sequence | Tikhonov´s deconvolution |
| Tomiyama et al.38 | 10 CAD patients |
Sequential 11.9 ± 8.78 days |
15O-water | Single bolus, single sequence | 1TCM |
| Engblom et al.24 | 21 patients with CAD |
Sequential 4–5 h between MRI and PET |
13N-ammonia | Single bolus, dual sequence | Distributed model |
| Kunze et al.25 | 29 patients with known or suspected CAD | Simultaneous | 13N-ammonia | Single bolus, dual sequence | Four different deconvolution methods |
| Kero 2019 (current work) | 12 patients with known or suspected CAD | Simultaneous | 15O-water | Single bolus, single sequence | 1TCM, direct estimation of PS and MBF |
| Study | MBF rest MRI and PET (mL/g/min) | MBF stress MRI and PET (mL/g/min) | Comparison of MRI-based and PET-based MBF (correlation and Bland-Altman comparison if available) | MFR MRI and PET | Comparison of MRI-based and PET-based MFR (correlation) |
|---|---|---|---|---|---|
| Pärkkä et al.22 |
0.71 ± 0.24 0.92 ± 0.26 |
1.72 ± 0.67 (K1) 3.76 ± 1.21 |
Regional MBF r = 0.80 |
2.51 ± 0.95 4.32 ± 1.78 |
Regional MFR r = 0.46 |
| Fritz-Hansen et al.20 |
0.80 ± 0.20 0.71 ± 0.16 |
1.83 ± 0.56 (K1) 2.03 ± 0.67 |
Regional MBF difference (stress-rest) r = 0.86 Regional MBF bias -0.28, LoA -0.45 – (-0.12) |
2.4 ± 0.8 2.9 ± 0.8 |
Global MFR r = 0.7 |
| Pack et al.37 |
1.03 ± 0.76 0.80 ± 0.24 |
2.97 ± 1.59 3.04 ± 1.14 |
Regional MBF r = 0.85 Regional MBF bias 0.12 |
3.2 ± 1.7 3.7 ± 0.7 |
n/a |
| Morton et al.21 | MBF values reported separately for different subject groups and coronary territories |
Regional MBF rest r = 0.32 Regional MBF stress r = 0.37 |
MFR values reported separately for different subject groups and coronary territories | Regional MFR r = 0.75 | |
| Qayyum et al.23 |
1.7 ± 0.49 0.75 ± 0.37 |
2.65 ± 1.77 1.90 ± 1.61 |
Global MBF difference (stress-rest) r = 0.81 Global MBF bias -0.11 ± 0.98 |
n/a |
Global MFR r = 0.89 Regional MFR r = 0.82 |
| Tomiyama et al.38 |
0.76 ± 0.10 0.71 ± 0.11 |
3.04 ± 0.82 3.09 ± 0.97 |
Global MBF r = 0.96 Regional MBF r = 0.92 |
4.13 ± 1.33 4.46 ± 1.43 |
Global MFR r = 0.93 Regional MFR r = 0.83 |
| Engblom et al.24 |
MRI rest and stress MBF range 0.6-3.8 PET rest and stress MBF range 0.6-4.0 |
Global MBF r = 0.92, rest bias 0 ± 0.2, stress bias -0.1 ± 0.5 Regional MBF r = 0.83, bias -0.1 ± 0.6 |
n/a |
Global MFR r = 0.69 Regional MFR r = 0.57 |
|
| Kunze et al.25 |
0.94 ± 0.23 0.78 ± 0.23 |
1.98 ± -0.49 1.89 ± 0.41 |
Slice average MBF r = 0.91 Regional MBF r = 0.84 |
2.09 2.53 |
Slice average MFR r = 0.60 |
| Kero 2019 (current work) |
0.97 ± 0.27 1.02 ± 0.28 |
3.19 ± 0.70 3.13 ± 1.16 |
Global MBF r = 0.86, rest bias 0.06, stress bias -0.05, stress LoA -1.58 – 1.71 Regional MBF r = 0.75 |
3.44 ± 0.97 3.05 ± 0.76 |
Global MFR r = 0.08, bias 0.39, LoA -1.94 – 2.73 |
1TCM, single-tissue compartment model; LoA, limits of agreement