Table 5.

The 11 nonlinear trimodal regression analysis parameters grouped by tissue type (fat, connective and muscle) were used to assess cardiovascular risks through machine learning algorithms and evaluation metrics were computed.

	Tissue	Algor.	Acc. Mean [%]	Acc. Max [%]	Sens. [%]	Spec. [%]	Recall [%]	Precision [%]	AUCROC
CHD	Fat	GB	73.8	75.2	69.6	78.0	69.6	75.9	0.828
		RF	79.6	82.2	76.1	83.1	76.1	81.8	0.884
		ADA-B	63.9	65.0	52.0	75.8	52.0	68.3	0.674
	Connective	GB	74.3	77.5	70.0	78.6	70.0	76.6	0.824
		RF	78.4	80.2	74.4	82.4	74.4	80.9	0.876
		ADA-B	63.3	65.4	56.3	70.5	56.2	65.6	0.680
	Muscle	GB	74.0	76.4	69.0	78.9	69.0	76.6	0.824
		RF	79.6	82.2	76.6	82.6	76.6	81.4	0.885
		ADA-B	63.6	66	63.3	63.9	63.3	63.7	0.673
CVD	Fat	GB	71.0	73.3	66.1	75.8	66.1	73.2	0.794
		RF	76.8	78.1	73.8	79.8	73.8	78.5	0.855
		ADA-B	61.5	64.1	50.8	72.2	50.8	64.7	0.645
	Connective	GB	71.3	74.3	66.0	76.7	66.0	73.9	0.792
		RF	76.1	78.5	71.8	80.3	71.8	78.5	0.846
		ADA-B	61.6	63.4	58.3	64.8	58.3	62.4	0.654
	Muscle	GB	70.2	72.8	65.0	75.5	65.0	72.6	0.788
		RF	76.8	78.9	73.7	79.9	73.7	78.6	0.854
		ADA-B	60.7	63.8	56.8	64.6	56.8	61.6	0.644
CHF	Fat	GB	83.0	85.0	80.2	85.9	80.2	85.0	0.918
		RF	88.4	90.0	87.3	89.4	87.3	89.2	0.956
		ADA-B	85.6	88.7	83.3	88.0	83.3	87.4	0.927
	Connective	GB	82.4	83.8	80.2	84.6	80.2	83.9	0.907
		RF	86.6	87.6	86.5	86.7	86.5	86.7	0.939
		ADA-B	82.9	85.2	80.9	84.9	80.9	84.3	0.905
	Muscle	GB	84.0	86.5	81.4	86.6	81.4	85.8	0.922
		RF	89.6	91.2	89.4	89.8	89.4	89.8	0.96
		ADA-B	87.4	89.1	85.7	89.2	85.7	88.8	0.943