Dose Reduction in Brain [¹⁸F]FDG PET/MRI: Give It Half a Chance

Integrated positron emission tomography (PET)/magnetic resonance imaging (MRI) offers promising tools for evaluating brain disorders, including the minimization of exposure to ionizing radiation. Considering the length of scanning time with PET/MRI systems and their high sensitivity, we assumed that the activity could be reduced by one half compared with recommended activity for brain 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) PET exams without degrading image quality.

We retrospectively simulated the reduction of injected activity (1 vs. 2 MBq/kg, [¹⁸F]FDG) in 100 patients assessed for cognitive impairment with simultaneous PET/MRI imaging. A list-mode acquisition was used to generate a 20-min image set as a reference (PET_STD) and to simulate a low-dose injection with a 10-min image (PET_LD). We tested the reproducibility between PET_LD and PET_STD with a blinded visual interpretation by two nuclear physicians asked to classify metabolic patterns, and a quantitative analysis conducted with regions-of-interest. Voxelwise comparisons between patients suggestive of Alzheimer's disease (AD) and frontotemporal dementia (FTD) were also conducted.

The intra-operator agreement was high between the PET_STD and PET_LD visual assessments for both readers (kappa 0.92 and 0.99). SUV ratios were strongly reproducible (intraclass correlation coefficient 0.95). The voxelwise and regional comparisons between AD vs. FTD metabolic profiles yielded very similar results with PET_STD and PET_LD.

A reduction of the [¹⁸F]FDG dose down to 1 MBq/kg is possible when performing 20-min brain PET/MRI without modifying diagnostic performance and quantitative assessments. The advantage is a significant reduction in the patient effective dose, which is non-negligible in longitudinal follow-up studies and in research protocols involving healthy volunteers.