Linearized and Kernelized Sparse Multitask Learning for Predicting Cognitive Outcomes in Alzheimer’s Disease

Table 3

Performance comparison of various methods with fusing multiple modalities data in terms of rMSE and nMSE on 10 cross validation cognitive prediction tasks.

Method	ADAS	MMSE	FLU	TRAILS
			ANIM	A	B
-MTL-MRI	6.494 ± 1.029	1.964 ± 0.306	4.911 ± 0.256	16.39 ± 2.906	55.82 ± 7.689
-MTL-PET	6.941 ± 1.244	2.118 ± 0.298	5.192 ± 0.145	16.56 ± 3.533	56.88 ± 9.447
-MTL-MP	6.219 ± 1.037	2.067 ± 0.293	4.928 ± 0.260	16.09 ± 2.768	53.70 ± 7.144
-MTL-ALL	6.174 ± 0.978	2.062 ± 0.272	4.789 ± 0.206	15.97 ± 2.785	53.37 ± 7.243
-MKMTL-MRI	6.369 ± 0.941	2.074 ± 0.291	4.993 ± 0.235	16.18 ± 3.089	55.95 ± 9.479
-MKMTL-PET	6.812 ± 1.155	2.060 ± 0.364	5.151 ± 0.227	16.61 ± 3.588	57.85 ± 11.24
-MKMTL-MP	6.112 ± 0.886	2.005 0.258	4.966 ± 0.269	16.13 ± 2.988	54.13 ± 9.450
-MKMTL-ALL	5.960 ± 0.834	1.959 ± 0.256	4.821 ± 0.224	16.00 ± 3.062	53.48 ± 9.592
-MKMTL-MRI	6.425 ± 0.951	1.951 ± 0.308	4.886 ± 0.264	16.11 ± 2.939	54.96 ± 7.499
-MKMTL-PET	6.783 ± 1.059	2.058 ± 0.323	5.107 ± 0.258	16.52 ± 3.515	55.51 ± 9.568
-MKMTL-MP	6.086 ± 0.987	1.917 ± 0.299	4.855 ± 0.249	15.95 ± 2.996	52.44 ± 8.074
-MKMTL-ALL	6.034 ± 0.978	1.905 ± 0.294	4.809 ± 0.244	15.88 3.028	52.20 ± 8.120
Method	RAVLT				nMSE
	TOTAL	TOT6	T30	RECOG
-MTL-MRI	10.18 ± 0.640	3.538 ± 0.147	3.735 ± 0.199	3.169 ± 0.306	10.24 ± 0.735
-MTL-PET	10.41 ± 0.441	3.627 ± 0.140	3.796 ± 0.176	3.258 ± 0.360	10.72 ± 1.163
-MTL-MP	10.01 ± 0.556	3.501 ± 0.149	3.693 ± 0.196	3.164 ± 0.314	9.710 ± 0.627
-MTL-ALL	9.755 ± 0.575	3.450 0.151	3.643 ± 0.200	3.172 ± 0.313	9.525 ± 0.608
-MKMTL-MRI	10.09 ± 0.605	3.532 ± 0.081	3.731 ± 0.253	3.203 ± 0.304	10.21 ± 1.019
-MKMTL-PET	10.30 ± 0.436	3.592 ± 0.145	3.754 ± 0.231	3.200 ± 0.357	10.82 ± 1.455
-MKMTL-MP	9.787 ± 0.375	3.471 ± 0.089	3.664 ± 0.199	3.159 ± 0.302	9.713 ± 0.968
-MKMTL-ALL	9.350 ± 0.460	3.402 ± 0.030	3.604 ± 0.221	3.196 ± 0.291	9.410 ± 0.985
-MKMTL-MRI	9.984 ± 0.525	3.477 ± 0.130	3.678 ± 0.204	3.143 ± 0.314	9.937 ± 0.753
-MKMTL-PET	10.19 ± 0.410	3.565 ± 0.146	3.745 ± 0.212	3.191 ± 0.351	10.31 ± 1.105
-MKMTL-MP	9.727 ± 0.467	3.397 ± 0.136	3.593 ± 0.162	3.112 ± 0.323	9.282 ± 0.869
-MKMTL-ALL	9.561 ± 0.442	3.361 ± 0.124	3.556 ± 0.170	3.104 ± 0.327	9.160 ± 0.860