| Low MRI signal in lungs |
Staging and delineation |
Reduced conspicuity |
Low proton density in lung parenchyma |
Hyper-polarised gas imaging, lower field strength (to increase relaxation times), or ultra-short echo time (UTE) sequences (less affected by fast T2 decay in lung parenchyma) |
| Motion during image acquisition |
Staging and delineation |
Motion artifacts |
Physiological motion (respiratory/cardiac) |
Acquisition with triggering or breath hold. Signal averaging, motion robust readouts |
| Poor visualisation of small airways on MR imaging |
Delineation and planning |
Potential hotspots (secondary to Lorentz force) that are not accounted for |
Bronchi are not well visualized due to short T2 |
Further development of ultra-short echo time sequences |
| Susceptibility induced field inhomogeneities |
Planning |
Reduced geometric fidelity |
Susceptibility differences at Air-Tissue interfaces |
Higher bandwidth, distortion corrections using B0 field maps, lower field strength |
| Synthetic CT generation difficult in thorax |
Planning |
Inaccurate results with current methods |
Short T2 of Lung tissue challenges current segmentation and contrast based approaches |
Continued research using specialised acquisition methods (e.g., ultrashort echo time) (29) |
| Lateral patient re-positioning limited |
Patient setup |
Less freedom in patient positioning |
Machine geometry |
Online re-planning to adapt to daily situation |
| Electron-return effect (Lorentz force) |
Planning |
Possible hotspots at air-tissue interface |
Altered path of secondary electrons when B>0 |
Accounted for within planning (30,31) |
| Motion during setup phase |
Verification |
‘Snapshot’ representation of setup image |
Physiological motion |
Align treatment position with setup position e.g., exhale imaging and gating. Or 4D-MRI acquisition with possibility for mid-position reconstruction (32-34) |
| Motion during treatment phase |
(treatment delivery) |
‘Intrafraction motion leads to dose ‘blurring’ necessitating increased RT planning margins’ |
Physiological motion |
Treatment on Mid-position, or implementation of gating/tracking |
| Motion during treatment phase |
(real-time imaging) |
Required temporal resolution too high for full volumetric cine imaging |
Physiological motion, inherent (lack of) speed in MRI acquisition |
Model based approaches that map volumetric information onto fast 2D acquisitions (35) |
