Table 2.

Metrics of models predicting brain age.

Model	Dataset size, $N$	Age range and distribution [range, mean ± std, (years)]	$MAE$ [mean ± std, (years)]	$R^2$
Dimitriadis and Salis (2017)	194	[18, 67], 37.7 ± 10.2	-	0.60 (EO) 0.48 (EC)
Zoubi et al. (2018)	468	[18, −], 34.8	6.87 ± 0.69	0.37 (EO)
Current model (MAE)	1,335	[5, 88], 38.8 ± 19.1	5.96 ± 0.33	0.81 (EO and EC)
Current model w/o “stacking tensor” and data augmentation	same		6.11 ± 0.50	0.80 (EO and EC)

The table shows the size of the datasets, mean age, and standard deviation (std), as well as the $MAE$(in years; lower is better) and R² (higher is better) metrics obtained on the corresponding datasets.