Article Figures & Data
Figures
- FIGURE 1.
Methodology for creating 18F-flortaucipir canonical images. (A) Creation of chronological 18F-flortaucipir dataset using associated 18F-florbetapir data. (B) Generation of 18F-flortaucipir K and NS canonical images from voxelwise modeling of chronological dataset.
- FIGURE 2.
Canonical images for 18F-flortaucipir PET derived from spatiotemporal modeling of chronological AD data in MNI152 space: tau canonical image (K) (A), nonspecific canonical image (NS) (B), and reference structural T1-weighted MR image (C).
- FIGURE 3.
TauIQ algorithm that estimates global and local tau outcome measures.
- FIGURE 4.
Examples of TauIQ decomposition of 18F-flortaucipir data into nonspecific, global, and local tau signals.
- FIGURE 5.
Cross-sectional analysis shown as box plots of distributions of 18F-flortaucipir TauL and SUVR outcome measures by clinical group.
- FIGURE 6.
Longitudinal analysis shown as box plots of distributions of change in 18F-flortaucipir TauL and SUVR outcome measures.
- FIGURE 7.
Correlations between clinical scores (CDR-SB, MMSE, and ADAS-Cog) and tau imaging outcome measures.
Tables
Outcome measure CN− vs. CN+ CN− vs. MCI+ CN− vs. dementia+ TauIQ TauL 1.00 (0.69–1.35), P < 0.10 1.53 (1.21–1.88) 2.70 (2.28–3.26), P < 0.44 SUVR Jack meta-ROI 0.69 (0.39–1.03) 1.58 (1.21–2.01) 2.64 (2.16–3.27) SUVR Braak I/II 0.55 (0.26–0.85) 1.62 (1.24–2.04), P < 0.64 2.46 (1.98–3.03) SUVR Braak III/IV 0.57 (0.28–0.88) 1.34 (0.99–1.72) 2.37 (1.93–2.92) SUVR Braak V/VI 0.49 (0.19–0.80) 0.97 (0.63–1.32) 1.90 (1.48–2.40) ↵* Effect sizes with CIs for 18F-flortaucipir TauL and SUVR outcome measures derived from group comparisons. Best-performing method in terms of effect size is highlighted in bold along with probability it is best method.
Parameter TauIQ TauL SUVR Jack meta-ROI SUVR Braak I/II SUVR Braak III/IV SUVR Braak V/VI CN− Mean and SD 1.20 (2.31) 0.003 (0.037) 0.001 (0.037) 0.003 (0.036) −0.001 (0.037) Effect size and CI 0.52 (0.28 to 0.79), P < 0.016 0.09 (−0.18 to 0.37) 0.02 (−0.25 to 0.30) 0.08 (−0.19 to 0.37) −0.02 (−0.30 to 0.24) n 940 33,812 557,745 37,422 847,562 CN+ Mean and SD 2.01 (2.97) 0.009 (0.044) 0.007 (0.039) 0.007 (0.041) 0.007 (0.039) Effect size and CI 0.68 (0.43 to 0.98), P < 0.022 0.21 (−0.12 to 0.55) 0.19 (−0.15 to 0.52) 0.18 (−0.16 to 0.55) 0.17 (−0.16 to 0.55) n 546 5,726 6,699 8,170 9,019 MCI− Mean and SD 1.87 (2.28) 0.008 (0.032) −0.002 (0.040) 0.004 (0.031) 0.010 (0.025) Effect size and CI 0.82 (0.57 to 1.58), P < 0.13 0.24 (−0.22 to 0.82) −0.04 (−0.49 to 0.53) 0.13 (−0.31 to 0.79) 0.40 (−0.05 to 0.94) n 377 4,458 190,970 14,512 1,610 MCI+ Mean and SD 3.61 (3.31) 0.023 (0.049) 0.008 (0.052) 0.023 (0.047) 0.017 (0.044) Effect size and CI 1.09 (0.79 to 1.54), P < 0.020 0.47 (0.12 to 0.85) 0.16 (−0.24 to 0.57) 0.49 (0.13 to 0.93) 0.39 (0.00 to 0.89) n 213 1,130 10,229 1,063 1,655 Dementia− Mean and SD 1.53 (2.43) −0.001 (0.032) −0.007 (0.019) 0.002 (0.024) 0.002 (0.024) Effect size and CI 0.63 (−0.13 to 2.14), P < 0.28 −0.04 (−1.55 to 0.76) −0.34 (−1.32 to 0.76) 0.09 (−1.18 to 1.01) 0.06 (−1.52 to 1.01) n 640 147,860 2,080 29,595 63,426 Dementia+ Mean and SD 2.52 (6.64) 0.004 (0.085) −0.011 (0.079) 0.002 (0.080) −0.005 (0.078) Effect size and CI 0.38 (−0.17 to 1.42), P < 0.27 0.05 (−0.44 to 0.83) −0.13 (−0.58 to 0.71) 0.02 (−0.56 to 0.76) −0.06 (−0.64 to 0.71) n 1,741 99,385 13,985 32,2607 67,106 ↵* Change in 18F-flortaucipir TauL and SUVR outcome measures per year for each clinical group. Best-performing method in terms of effect size (mean/SD) is highlighted in bold along with probability that it is the best method. n is number of subjects, in both active and placebo arms of simulated clinical trial, required to show 25% reduction in tau accumulation in active arm over period of 1 y (power = 80%, α = 0.05).
Supplemental Data
Files in this Data Supplement:
