Skip to main content
. 2017 Jul 21;44(9):4665–4676. doi: 10.1002/mp.12416

Table 2.

Summary of runs test results and AIC and BIC scores for model fits to 30 randomly selected voxel TACs. Lowest AIC and BIC scores are underlined, indicating the best model according to that measure. AIC and BIC values have been calculated using both the scaled and unscaled versions as indicated in the table

Model 2C3K 2C4K 3C5K 3C6K
Runs test passes from fits to all 30 TACs
Runs passes 22 27 30 30
Information criteria summed for all statistical simulation TACs
AIC (scaled) 1793 1500 1424 1508
BIC (scaled) 2002 1738 1687 1790
AIC (unscaled) 2826 2595 2473 2549
BIC (unscaled) 2999 2801 2702 2801
Numbers of TACs for which each model has the lowest scores
AIC (scaled) 4 8 18 0
BIC (scaled) 8 7 15 0
AIC (unscaled) 4 6 20 0
BIC (unscaled) 8 7 15 0