Table 6.

Elastic Net Regression performance – retrospective data merging scenario – Summary of the results in terms of variance explained by the model and by age and scanner, the features considered most relevant in our study. Full details of all the other features are provided in Supplementary Table S3.

		Scanner upgrade scenario					Retrospective data merging scenario
		Analysis option A	Analysis option B	Analysis option C	Analysis option D	Analysis option E	Analysis option A	Analysis option B
Variance explained by the model		0.243	0.161	0.207	0.173	0.125	0.190	0.098
Variance explained by the features	Age	0.060	0.043	0.048	0.054	0.034	0.052	0.070
	Scanner	0.046	0.012	0.066	0.008	0.000	0.115	0.000

Options tested in our study are: (I) for the scanner upgrade scenario: (A) without BC, single-site training, FA included; (B) with BC, single-site training, FA included; (C) with BC, site-specific training, FA included; (D) with BC, mixed training, FA included; (E) with BC, mixed training, FA excluded; (II) for the retrospective data merging scenario: (A) with BC, site-specific training, FA excluded; (B) with BC, mixed training, FA excluded. The amount of WMH variance explained by the model is calculated using the R-squared coefficient and reported in the first row. The amount of WMH variance explained by the features is reported in the rest of the table for the most relevant variables (age and scanner).