Table 4. Absolute and percentage difference and root mean squared error (RMSE) between methods used to compute single kidney volume (SKV) by the expert operator on MR and CT images from ADPKD patients in the experimental dataset.

	SKV computation method	SKV Difference ml	RMSE ml	SKV Difference %	RMSE %
MR	Osirix free-hand vs ImageJ polyline	-19 [–229, 60]	66	-0.8 [-9.5, 4.4]	3.2
	Livewire tool vs ImageJ polyline	-10 [–123, 93]	40	-1.4 [-15.9, 5.4]	4.0
	Stereology vs ImageJ polyline	-47 [–329, 152]	101	-3.7 [-23.1, 5.2]	6.3
	Mid-slice vs ImageJ polyline	-19 [–251, 346]	118	-2.9 [-32.1, 14.4]	9.7
	Ellipsoid vs ImageJ polyline	-213 [–1044, 333]	350	-18.8 [-48.0, 53.2]	25.4
CT	Osirix free-hand vs ImageJ polyline	2 [–40, 82]	23	-0.1 [-2.7, 3.2]	1.3
	Livewire tool vs ImageJ polyline	-26 [–88, 42]	39	-2.2 [-6.2, 1.6]	2.8
	Stereology vs ImageJ polyline	-11 [–67, 39]	26	-1.0 [-5.2, 3.2]	2.1
	Mid-slice vs ImageJ polyline	-190 [–862, 23]	276	-13.4 [-27.1, 2.8]	15.0
	Ellipsoid vs ImageJ polyline	-128 [–666, 559]	289	-10.7 [-33.6, 34.5]	19.0

Number of single kidneys analyzed n = 30 for MR and n = 30 for CT. SKV are from first tracing of expert operator.

SKV difference is expressed as mean difference and [range]; RMSE = Root mean square error.