Table 2:

OMFA vs transfer learning for various MRI model predictors^{^a}

Predictors for Model	OMFA, LOOCV (r), and MAE		TL, LOOCV (r), and MAE
rCBV	0.27	17.79	0.53	15.19
rCBV, EPI + C	0.25	18.03	0.63	11.65
rCBV, FA	0.34	17.24	0.58	11.31
rCBV, MD	0.28	17.74	0.60	11.93
rCBV, T1 + C	0.33	16.69	0.69	11.15
rCBV, T2WI	0.26	17.96	0.59	12.30
rCBV, FA, MD	0.32	17.47	0.66	11.93
rCBV, T1 + C, T2WI	0.35	16.61	0.75	9.03
rCBV, T1 + C, FA	0.39	16.55	0.73	9.07
rCBV, T1 + C, MD	0.35	16.77	0.74	9.41
rCBV, T2WI, FA	0.32	17.47	0.64	10.94
rCBV, T2WI, MD	0.26	18.02	0.64	11.15
rCBV, T1 + C, FA, MD	0.37	16.78	0.85	6.73
rCBV, T2WI, FA, MD	0.30	17.68	0.69	10.95
rCBV, T1 + C, T2WI, FA	0.37	16.79	0.73	9.41
rCBV, T1 + C, T2WI, MD	0.34	16.88	0.78	7.01
rCBV, T1 + C, T2WI, FA, MD	0.35	17.05	0.88	5.66
rCBV, T1 + C, T2WI, FA, MD, EPI + C	0.34	17.17	0.86	6.09

Note:—MAE indicates mean absolute error.
↵a OMFA models were generated on the basis of linear regression analysis. Both (r) and MAE were determined using LOOCV to plot model-predicted TCD against actual TCD from spatially matched biopsies (n = 82).