To compare myelin water fraction (MWF) and myelin water content (MWC) in new lesions of subjects with multiple sclerosis (MS).Myelin water fraction is a histopathologically validated marker for myelin¹ which has become a useful technique for measuring changes in myelination in vivo. MWF is defined as the fraction of water with a short T₂ component corresponding to the water trapped between the myelin bilayers (labelled myelin water) relative to the total water.² Therefore, increases in total water content (WC) of tissue would decrease the MWF and mimic demyelination. This problem is particularly relevant in new MS lesions which may have demyelination but also show significant increases in WC at first appearance, related to edema and inflammation, but resolve at later times. Our group has recently optimised a method for determining total water content in the brain.^3,4 This allows us to now make corrections to the myelin water fraction and calculate a myelin water content (MWC). By comparing MWF to MWC it would be possible to assess if changes in MWF are due to increases in WC or true decreases in myelin.6 subjects with relapsing remitting MS (mean age=33y, median EDSS=3, mean disease duration=6y), were scanned on a 3.0T Philips Achieva system (Best, The Netherlands) every month for 6 months. Scans included a 32 echo spin-echo sequence (TE/TR=10/1200ms, 1x1x5mm³, 7 slices),⁵ an inversion recovery MPRAGE sequence (TR/TE=6.5/3.2ms, TI=150,400,750,1500,3500ms, 1x1x5mm³, 13 slices), a 3D-T1 turbo field echo sequence (TR/TE=28/4ms) and a FLAIR sequence (TR/TI/TE=900/2500/80ms). Voxel-wise T₂ decay curves were analyzed using non-negative least squares with stimulated echo correction⁶ and spatial regularisation⁷. MWF was calculated from T₂ distributions as the contribution between 15 and 40 ms divided by the total T₂ signal. Water content was calculated using the cerebrospinal fluid reference method outlined in Figure 1.^3,4 All MWF and WC images were registered to the month 0 time-point for each person using FLIRT (from the FSL toolbox). New lesions were delineated at the time of first appearance and propagated to all other time-points. These lesion areas were then overlaid on the registered MWF and WC images to extract the mean for each region of interest. MWC was defined as MWF × WC.23 new lesions were found in the 6 subjects. Plots of MWF and MWC for a sampling of new lesions are shown in Figure 2. A very strong correlation between MWF and MWC was observed (R²=0.96, p<0.0001) (Figure 3). The shape of the MWF and MWC curves over time showed great similarity with the MWC curve having smaller values. On average, MWC was 22% lower than MWF. Changes in WC do not appear to drive the changes in MWF seen in new lesions. The shapes of the MWF and MWC curves were similar and the correlation between MWF and MWC was very strong. This lack of difference between MWF and MWC is not unexpected since even large water content changes would only translate to small changes in MWF.⁸ As with previous studies,⁹ differences between individual lesions are evident suggesting that MWF can be used to monitor pathological changes. Measuring WC involves additional steps beyond simply examining MWF; while correcting MWF for WC to obtain MWC is valuable for examining changes around lesion first appearance, MWF is still a valid measure of change after first appearance and can be used to investigate differences between lesions and lesion recovery.Conclusion: MWF can be used to characterise and monitor chronic lesions as MWF values mirror MWC values. Changes around the time of appearance can be explored in greater depth by including a measure of WC so that edema/inflammation can be distinguished from demyelination.