Skip to main content
. 2023 Nov 13;9(4):20552173231211396. doi: 10.1177/20552173231211396

Figure 5.

Figure 5.

Evaluation of reproducibility for CSF MTR, WM MTR, CSF PSR, and WM PSR in healthy control subjects. Plots demonstrate Bland–Altman characterization of reproducibility, where red-dotted lines indicate 95% confidence interval of mean difference, black solid line shows the mean difference. In each plot, the 95th percentile overlaps 0, indicating no observable bias. CSF MTR and PSR show strong agreement as shown by ICC, with the latter showing slightly greater reproducibility compared to the former, however, the mean coefficient of variation for the former is lower than the latter. For WM, the ICC for MTR is clustered and does not show a strong correlation with follow-up, however, the coefficient of variation for WM MTR is low, as is WM PSR. It is expected that the variation in WM is lower as both the MTR and PSR are designed to be sensitive to white matter changes, more so than tailored for CSF evaluation.

MTR: magnetization transfer ratio; PSR: pool size ratio; CSF: cerebrospinal fluid; WM: white matter.