170 Part II Chapter 6 tendency’ of the parameter values, as outliers influence this statistical measure. Also ‘interquartile range’ might suit the description of an asymmetrical distribution better than the single value ‘standard deviation’, which assumes a more normal distribution. In a standardized normal distribution, mean+3SDwould define the upper 0.1% of voxel values within a VOI, which would result in (reproducible) selection of voxels visually pathologically increased (mean+2SD would lead to a substantial part of normal bone to be included in the FD VOIs). In each of the 5 VOIs however, the number of voxels was 8. Therefore, based on the actual distribution it was only possible the find a cutoff for the upper 12.5% (i.e. the hottest voxel within a single ROI). Practically, we are not aware of any software available that provides us with other statistical measures of a VOI (e.g. median, p25 and p75 or mean/SD from log-transformed voxel values), and Rohren, et al. and Lapa, et al. used a similar method with SUVmean values [6, 7]. Another limitation of the current study is that only a small and heterogeneous cohort selected for either start of bone remodeling therapy or planned surgery was available for analysis, mostly inherent to FD being a rare disease. Other disadvantages of Na[18F]F-PET/CT for daily routine compared to (planar) bone scintigraphy, are the expenses and availability of both equipment and tracer, compared to planar bone scintigraphy. At this cost, Na[18F]F-PET/CT provides superior resolution and much faster acquisition when compared to whole body quantitative SPECT/CT. Radiation burden is currently equivalent for Na[18F]F-PET/CT and 99mTc-HDP planar scintigraphy, with a median radiation burden of 1.7 mSv for ~1.0 MBq/kg Na18F and 2-4 mSv for the total body low dose CT versus 3.1 mSv for 550 MBq 99mTc-HDP and, if indicated, 0.2-2 mSv per SPECT-bed position low-dose CT depending on the region [27-29]. In conclusion, we were able to reproducibly quantify normal bone activity and baseline FD disease activity on Na[18F]F-PET/CT using a personalized approach, in a reliable and clinically feasible manner. Furthermore, when compared to SBS (the current standard), this resulted in 3D, improved quantitative parameters obtained with Na[18F] F-PET/CT, exploiting the superior spatial resolution of PET/CT over SBS. Na[18F]F- PET/CT using TLF potentially provides new clinical insight in baseline FD disease burden, as this parameter showed high correlation with markers of bone formation. Therefore, Na[18F]F-PET/CT could be a meaningful tool to evaluate treatment efficacy during long-term follow-up in FD patients. SBS is known to remain stationary, irrespective of use of medication, whereas TLF on Na[18F]F-PET/CT was higher in baseline patients using bisphosphonates, which implies that Na[18F]F-PET/CT is a promising modality to quantitatively measure treatment efficacy [2].