AbstractBackgroundRegional patterns of brain atrophy are characteristic features of neurodegenerative disorders. Radiological assessments of cortical brain atrophy tend to reflect sulcal widening and brain shrinkage, whereas quantitative assessments based on grey matter volume or cortical thickness can be hampered by poor SNR and inaccurate estimation of grey/white matter boundaries. Our aim was to define a simple and rapidly‐computable marker of localized sulcal widening based on standard brain tissue segmentation tools and assess it in a cohort of individuals with frontotemporal dementia (FTD) and SV2A PET scans.Method3DT1 MRI and [11C]UCB‐J PET scans were analyzed from individuals with FTD (N=8, MMSE range 10‐28) and age‐matched healthy volunteers (HV, N=17) within the MIND‐MAPS consortium. 3DT1 images were segmented into gray matter (GM), white matter (WM), and cerebral spinal fluid (CSF) using SPM12. The lobar CSF occupancy fraction (CSFOF) was calculated as CSF/(GM+WM+CSF) in the frontal, parietal, temporal and occipital lobes. The tissue type masks were created by a threshold of total tissue occupancy > 50% then assigning each voxel to the tissue for which it has the maximum probability. 3DT1 scans were also assessed using cortical thickness (Freesurfer v6.0). SV2A binding was quantified as distribution volume ratio (DVR‐1) from [11C]UCB‐J using centrum semiovale as reference region.ResultThe CSFOF computation time was approximately 6 minutes per scan on a 2019 MacBook pro. CSFOF was significantly higher in FTD than in HV (ANOVA F=19.6, p<0.0001, Figure 1). Corrected, non‐parametric linear relationships between CSFOF and MMSE, cortical thickness and [11C]UCB‐J DVR‐1 are summarized in Table 1 and Figure 2 for the frontal lobe. The relationships were weaker in the FTD cohort alone, compared with the full sample. Cortical thickness was the only nominally significant association remained after correction for age and MMSE, suggesting that CSFOF and SV2A density may reflect different aspects of the disease.ConclusionThe CSFOF provides a simple, rapid and robust quantification of lobar brain atrophy, reflecting sulcal widening and both grey and white matter atrophy. It complements other imaging markers of neurodegeneration and extends the concepts of ventricular volume and hippocampal occupancy fraction to the wider neocortex.