Computed tomography volumetry vs three-dimensional surface scanning for liver graft size estimation using explant liver mass

Table 1 Patient demographics, median (interquartile range)

Parameters	Study group, n = 19
Sex
Male	13
Female	6
Age in years	54.0 (40.0-63.5)
Height in cm	172.0 (163.0-181.0)
Weight in kg	80.0 (65.5-96.0)
BMI in kg/m2	26.3 (22.7-32.0)
Thoracic circumference	93.0 (85.0-103.5)
Abdominal circumference	97.0 (86.5-107.5)

BMI: Body mass index.

Table 2 Calibration model of 3D surface volumetry against explant liver mass

Variable	β coefficient (95%CI)	SE	P value
Intercept	299.7 (-129.40 to 728.77)	203.38	0.16
3D surface estimate	1.01 (0.73 to 1.29)	0.13	< 0.001

CI: Confidence interval; 3D: Three-dimensional.

Table 3 Calibration model of computed tomography volumetry against explant liver mass

Variable	β coefficient (95%CI)	SE	P value
Intercept	221.86 (-157.08 to 600.79)	179.60	0.234
CT volumetry	1.04 (0.80 to 1.28)	0.11	< 0.001

CT: Computed tomography; CI: Confidence interval.

Table 4 Accuracy of graft size estimation relative to explant liver mass, n = 19

Method	Median absolute error (%)	IQR (%)
CT volumetry	18.45	9.72
3D surface-derived volumetry	21.26	12.87
Paired comparison (3D-CT)
Statistic	Value
Median paired difference (%)	2.44
95%CI	0.00-3.9
Wilcoxon signed-rank P value	0.011

IQR: Interquartile range; CI: Confidence interval; CT: Computed tomography; 3D: Three-dimensional.

Table 5 Intraclass correlation coefficients for agreement between imaging-based graft size estimates and explant liver mass

Method	ICC (3, 1)	95%CI	n
CT volumetry vs explant liver mass	0.90	0.76-0.96	19
3D surface-derived volumetry vs explant liver mass	0.87	0.70-0.95	19

CI: Confidence interval; CT: Computed tomography; ICC: Intraclass correlation coefficient; (3, 1): Two-way mixed-effect model; n: Number of liver explants; 3D: Three-dimensional.