Table 2 Acquisition technique recommendations for PWI in HNSCC

1. DCE perfusion, achieved with 3D Spoiled-Gradient Echo T1-weighted sequences, should be used

2. Limited evidence exist on the applicability of DSC and ASL perfusion techniques in the clinical setting

DCE acquisition parameters

1. Axial images should be acquired on scanners with a field strength of 1.5 or 3.0 Tesla using head and neck or neurovascular coils, with the following minimum parameters

i. Slice thickness: ≤ 4 mm (recommended: 3 mm)

ii. Gap: no gap

iii. In-plane resolution: ≤ 2 × 2 mm

iv. Fat suppression

2. T1 mapping sequences with multiple flip angles (suggested not inferior to 3, ideally ranging from 5° to 30°) should be acquired, with the same geometry of DCE, prior to contrast administration to obtain accurate data kinetic fitting

3. Temporal resolution should be ≤ 5 s, with 5 acquisitions prior to contrast agent administration and a total acquisition time of at least 5 min

Contrast agent

1. Gadolinium-based contrast agents should be employed at the recommended dosage based on molecule choice with an injection rate ≥ 2 mL/s, followed by at least 20 mL saline flush

Image analysis

1. Bi-compartimental and extended Tofts models are the current reference standard for quantitative perfusion parameter calculation

2. The arterial input function should be measured placing a ROI in the carotid arteries (external/internal or common based on lesion location and best time-intensity curve morphology). The use of a population-based AIF can be considered

Quantitative DCE parameters calculation

1. ROIs should be placed on DCE post-contrast images avoiding areas of hemorrhage, necrosis, cystic components and neighboring vessels and pasted on perfusion maps to extract the quantitative parameters