Direct prospective comparison of <sup>18</sup>F-FDG PET and arterial spin labelling MR using simultaneous PET/MR in patients referred for diagnosis of dementia.
prospective_cohort · Level II
Where this comes from
- Record sourced from PubMed, PMID 31960098.
- Also identified by DOI 10.1007/s00259-020-04694-1.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
<sup>18</sup>F-FDG PET is routinely used as an imaging marker in the early and differential diagnosis of dementing disorders and has incremental value over the clinical neurological and neuropsychological evaluation. Perfusion MR imaging by means of arterial spin labelling (ASL) is an alternative modality to indirectly measure neuronal functioning and could be used as complement measurement in a single MR session in the workup of dementia. Using simultaneous PET-MR, we performed a direct head-to-head comparison between enhanced multiplane tagging ASL (eASL) and <sup>18</sup>F-FDG PET in a true clinical context of subjects referred for suspicion of neurodegenerative dementia. Twenty-seven patients underwent a 20-min <sup>18</sup>F-FDG PET/MR and simultaneously acquired eASL on a GE Signa PET/MR. Data were compared with 30 screened age- and gender-matched healthy controls. Both integral eASL and <sup>18</sup>F-FDG datasets were analysed visually by two readers unaware of the final clinical diagnosis, either in normal/abnormal classes, or full differential diagnosis (normal, Alzheimer type dementia [AD], dementia with Lewy Bodies [LBD], frontotemporal dementia [FTD] or other). Reader confidence was assessed with a rating scale (range 1-4). Data were also analysed semiquantitatively by VOI and voxel-based analyses. The ground truth diagnosis for the patient group resulted in 14 patients with a neurodegenerative cognitive disorder (AD, FTD, LBD) and 13 patients with no arguments for an underlying neurodegenerative cause. Visual analysis resulted in equal specificity (0.70) for differentiating normal and abnormal cases between the two modalities, but in a higher sensitivity (0.93), confidence rating (0.64) and interobserver agreement for <sup>18</sup>F-FDG PET compared with eASL. The same was true for assigning a specific differential diagnosis (sensitivity: and 0.39 for <sup>18</sup>F-FDG PET and eASL, respectively). Semiquantitative analyses revealed prototypical patterns for AD and FTD, with both higher volumes of abnormality and intensity differences on <sup>18</sup>F-FDG PET. In a direct head-to-head comparison on a simultaneous GE Signa PET/MR, <sup>18</sup>F-FDG PET performed better compared with ASL in terms of sensitivity and reader confidence, as well as volume and intensity of abnormalities. However, using pure semiquantitative analysis, similar diagnostic accuracy between the two modalities was obtained. Therefore, ASL may still serve as complement to neuroreceptor or protein deposition PET studies when a single simultaneous investigation is warranted.
Medical subject headings
- Alzheimer Disease
- Fluorodeoxyglucose F18