Table 12.

Comparison among the features used in exist state-of-the-art works and different combinations of the proposed features on ICU admission prediction on Site C dataset.

Features	AUC			Sensitivity (PPV = 70%)			K
	Mean	95% CI	p value	Mean	95% CI	p value
Img feature (Tang 2020)	0.763	(0.670, 0.856)	$p=0.044$	85.0%	(76.4%, 93.6%)	$p=0.020$	10
Img feature (Zhu 2020)	0.675	(0.645, 0.706)	$p<0.001$	73.9%	(59.0%, 88.8%)	$p<0.011$	39
DVB	0.595	(0.524, 0.665)	$p<0.001$	63.3%	(36.1%, 90.5%)	$p=0.019$	4
HLQ	0.691	(0.660, 0.722)	$p<0.001$	86.7%	(80.3%, 93.0%)	$p<0.014$	7
WLR	0.815	(0.782, 0.848)	$p=0.020$	95.6%	(92.8%, 98.3%)	p = 0.186	12
WLR + HLQ	0.826	(0.813, 0.839)	p = 0.191	96.1%	(94.4%, 97.8%)	–	20
WLR+DVB	0.835	(0.809, 0.861)	p = 0.365	95.0%	(91.6%, 98.4%)	p = 0.088	15
HLQ + DVB	0.760	(0.705, 0.815)	$p=0.016$	85.6%	(77.6%, 93.5%)	$p<0.010$	2
WLR + HLQ + DVB	0.840	(0.804, 0.876)	–	94.4%	(92.3%, 96.6%)	$p=0.035$	35