Recommendations on Patient Preparation, Acquisition Protocols, and Image Reconstruction
| Category | Recommendation |
| Patient preparation | |
| Fasting period | Minimum of 6 h before scanning, with plain water allowed |
| Blood glucose level | Minimum of 4 mmol/L, with upper limit determined by study objectives |
| Patient positioning | Arms above head when possible, with maximum body mass index of 35 kg/m2 allowed for consistent arm positioning |
| Administered 18F-FDG activity and injection duration | Activity no more than needed for sufficient image quality to meet study objectives, with injection duration documented and kept consistent for all patients |
| Acquisition protocol | |
| Static vs. dynamic | Static scans acceptable if compatible with study objectives; dynamic scans recommended when tracer quantification for specific lung compartments is needed; rationale for chosen acquisition method reported |
| Respiratory gating | End-expiration gating results reported when used; list-mode data stored to allow future reprocessing as techniques improve |
| Accounting for respiratory motion | Breath-hold at end-expiration or mid-expiration most frequently used to match lung volumes for CT and PET; prescan coaching of breathing instructions recommended (coaching patient on breathing instructions for attenuation correction CT can minimize most respiratory motion errors); exploration of approaches such as cine-CT |
| Duration and time frame | Static imaging: starting at 60 min after injection |
| Dynamic imaging: 45- to 90-min acquisition starting immediately after injection; 60-min scan is routinely tolerated (breaks may be required for improved patient tolerance) | |
| Time frames for typical 60-min dynamic acquisition (time after injection): 0–2 min: 5–15 s/frame; 2–5 min: 20–30 s/frame; 5–10 min: 60 s/frame; 10–18 min: 120 s/frame; 18–30 min: 180 s/frame; 30–60 min: 300 s/frame | |
| Image reconstruction | Method harmonized as much as possible in multicenter study; for iterative reconstruction, larger number of iterations should be considered to ensure uniform convergence, regardless of image reconstruction algorithm, followed by suitable filter to control noise if desired |