• David Fällmar, Johan Lilja, Vilma Velickaite, Torsten Danfors, Mark Lubberink, André Ahlgren, Matthias J P van Osch, Lena Kilander, and Elna-Marie Larsson.
• Department of Surgical Sciences/Radiology, Uppsala University, Uppsala, Sweden.
• J Neuroimaging. 2016 May 1; 26 (3): 324-30.

PurposeFunctional imaging is becoming increasingly important for the detection of neurodegenerative disorders. Perfusion MRI with arterial spin labeling (ASL) has been reported to provide promising diagnostic possibilities but is not yet widely used in routine clinical work. The aim of this study was to compare, in a clinical setting, the visual assessment of subtracted ASL CBF maps with and without additional smoothing, to FDG-PET data.MethodsTen patients with a clinical diagnosis of dementia and 11 age-matched cognitively healthy controls were examined with pseudo-continuous ASL (pCASL) and 18F-Fluorodeoxyglucose positron emission tomography (FDG-PET). Three diagnostic physicians visually assessed the pCASL maps after subtraction only, and after postprocessing using Gaussian smoothing and GLM-based beta estimate functions. The assessment scores were compared to FDG PET values. Furthermore, the ability to discriminate patients from healthy elderly controls was assessed.ResultsSmoothing improved the correlation between visually assessed regional ASL perfusion scores and the FDG PET SUV-r values from the corresponding regions. However, subtracted pCASL maps discriminated patients from healthy controls better than smoothed maps. Smoothing increased the number of false-positive patient identifications. Application of beta estimate functions had only a marginal effect.ConclusionSpatial smoothing of ASL images increased false positive results in the discrimination of hypoperfusion conditions from healthy elderly. It also decreased interreader agreement. However, regional characterization and subjective perception of image quality was improved.Copyright © 2015 by the American Society of Neuroimaging.

28 keywords...

hide…