Table 3.

Association of TAI measures with outcomes using logistic regression analysis receiver operating curve analysis, linear regression and Akaike and Bayesian information criterion

	Clinical outcome					Admission deficit
	Univariate logistic regression analysis					Linear regression
10 HU increase of TAI	OR (95% CI)	p	AUC	AIC	BIC	B (95% CI)	p	AIC	BIC
Arterial phase	1.21(1.041–1.414)	0.01	0.62	296.1	302.9	−0.87(−1.4 to − 0.34)	0.002	903.0	909.7
Venous phase	1.148(1.011–1.303)	0.03	0.58	297.8	304.6	−0.33(−0.78 to 0.11)	0.139	910.9	917.7
Delayed phase	1.046(0.919–1.19)	0.50	0.53	302.3	309.1	−0.10(−0.59 to 0.39)	0.681	913.0	919.7
TiCTA	1.153(1.021–1.301)	0.02	0.60	297.0	303.8	−0.44(−0.87 to − 0.01)	0.044	909.0	915.8

TAI thrombus attenuation increase, OR odds ratio, aOR adjusted odds ratio, CI confidence interval, B linear regression coeficient, AUC area under the curve, AIC Akaike information criterion, BIC Bayesian information criterion, p significance value, HU Hounsfield unit, TiCTA time invariant computed tomography angiography; and italic p is significant at the 0.05 level