Table 3.

The performance of deep learning framework

	Sensitivity %	Specificity %	AUC	p value
POI	92.41 (73 of 79) [84.85, 97.50]	90.49 (409 of 452) [87.60, 92.79]	0.97 [0.95, 0.98]	< 0.001
iHU	91.14 (72 of 79) [82.44, 96.15]	41.59 (188 of 452) [36.95, 46.26]	0.69 [0.63, 0.74]	< 0.001
POI + iHU	93.67 (74 of 79) [86.60, 97.73]	88.05 (398 of 452) [85.07, 90.81]	0.97 [0.94, 0.98]	< 0.001

Note: Values in parentheses are the numbers for the percentage calculation. Values in brackets are 95% confidence intervals [95%CI, %]. AUC, area under the receiver operating characteristic curve; POI, portion of infection; iHU, average infection Hounsfield unit