Optimization of brain perfusion image quality by cortical surface-based projection of arterial spin labeling maps in early-onset Alzheimer's disease patients.

OBJECTIVES: Arterial spin labelling (ASL) is a promising MRI sequence that allows noninvasive detection of cortical perfusion alterations in neurodegenerative disorders, but its interpretation remains difficult at an individual level. In this work, a cortical surface-based projection of ASL maps was applied in patients with early-onset Alzheimer's disease (EOAD) to improve the image quality and visual representation of perfusion data.
METHODS: Eighteen patients referred from the reference centre for EOAD were assessed by MRI with ASL sequences. Data processing was applied on each examination including correction of partial volume effects and cortical projection of preprocessed ASL data. Cortical segmentation and perfusion display were qualitatively analyzed according to a three-point scale.
RESULTS: All examinations were suitable for complete data processing. Quality of segmentation and of cortical surface-based perfusion maps was scored as optimal in 72 % in both cases. Cortical surface-based ASL maps provided a more global view than single slices and an accurate approach of brain perfusion in EOAD patients.
CONCLUSION: Cortical surface-based analysis of ASL maps is technically feasible with a good image quality and may enable significant improvement in the detection of focal perfusion alterations in neurodegenerative disorders in the real-life clinical setting. KEY POINTS: • Arterial spin labelling is a promising sequence for assessing Alzheimer's disease. • Optimization of ASL brain perfusion image quality is crucial for image interpretation. • Cortical surface-based analysis may improve detection of perfusion alterations in a real-life clinical setting.