17. Mean (COV) and Positivity Threshold* * Positivity threshold calculated from the 100 controls (of 138) with lowest mean cortical amyloid values. Threshold = region mean + 3 SD
38. Group average images AV – negative (n=12) row 1 AV – positive (n=17) row 2 PiB – negative (n=12) row 3 PiB – positive (n=17) row 4
39. Group average surface maps AV – negative row 1 AV – positive row 2 PiB – negative row 3 PiB – positive row 4
40. The relative values of cerebellar gray, pons and white matter are similar for PiB and AV45 (slopes ~1.0 across these structures). Slopes for cortical regions are ~0.60-0.65 (for all reference tissue normalizations)
Subjects plots in order of Subject ID xxx_S_yyyy. Original ADNI Reference of Atlas cereb gray voi
Note the two sections of correlated cortex values with Pons and White Matter. Caused by variability in cerebellar gray values. Likely negative if cortex is markedly lower than pons or white matter. Likely positive if cortex similar to pons, and nearing white matter.
Same for controls, except fewer positive scans. Also white matter values a little higher, hence cortex normalized to white matter will be lower.
Note tighter distribution of alternative ref regions using combined region, Also best distinction of cortex
Threshold defined as mean + 3 std devs of the lowest 100 controls
Point out: 1) tresholds highly dependent on reference , slight dependence on target. 2) Combined more suitable because the amyloid “negative” cortical values have lower variance !
Note effect of lower white matter relative to cerebellum/pons in MCI, causes cortical values in amyloid negative to be slightly higher.
Regions show similar sensitivity except for occipital (least positive)
Note white matter has most subjects positive. Remember that white matter is lower in MCI relative to cerebellum and pons.
Mostly scatter correction effects: which effects image contrast: HRRT old highest, then BioGraph, then GE, then Phlips, then HR+ then HRRT new Effects reduced by white matter or combined normalization
Standardized image scale. If all references regions worked perfectly, then images using different reference regions would be scaled and hence look exactly the same.
Just the n=17
Note negative subjects haven’t changed much, as PVC effects CBL and CTX about the same. In amyloid positive subjects, get considerable enhancement