Comparison of 3T MR scanners in regional cartilage-thickness analysis in osteoarthritis: a cross-sectional multicenter, multivendor study

Table 2 Results of Philips intrascanner and vendor interscanner comparisons

Scanner pair	Region	RMS COV	Mean difference/bias (mm)	95% limits of agreement (mm)		Paired ttest
				Lower bound	Upper bound	Pvalue
Philips first to second scan(intrascanner)	cLF	2.6%	0.00	-0.19	0.20	0.95
	cMF	3.2%	0.03	-0.14	0.20	0.23
	LT	4.1%	-0.05	-0.34	0.25	0.30
	MT	4.6%	0.00	-0.30	0.30	0.98
	LF	3.6%	0.00	-0.23	0.23	0.98
	MF	2.8%	0.05	-0.08	0.18	0.02^a
Philips to Siemens	cLF	4.1%	0.06	-0.13	0.25	0.10
	cMF	5.4%	0.05	-0.19	0.30	0.04^a
	LT	4.0%	0.06	-0.10	0.22	0.75
	MT	5.6%	0.05	-0.33	0.43	0.30
	LF	4.9%	0.13	-0.04	0.30	0.00^a
	MF	5.8%	0.08	-0.12	0.29	0.00^a
Siemens to GE	cLF	5.1%	0.12	-0.11	0.36	0.00^a
	cMF	5.6%	0.13	-0.06	0.31	0.00^a
	LT	7.7%	-0.24	-0.55	0.06	0.00^a
	MT	7.7%	-0.17	-0.51	0.18	0.01^a
	LF	4.2%	0.04	-0.20	0.27	0.28
	MF	5.2%	0.01	-0.25	0.26	0.85
Philips to GE	cLF	6.6%	0.18	-0.03	0.39	0.00^a
	cMF	9.6%	0.21	-0.03	0.46	0.00^a
	LT	8.1%	-0.23	-0.60	0.13	0.00^a
	MT	5.4%	-0.12	-0.34	0.11	0.00^a
	LF	7.2%	0.17	-0.14	0.47	0.00^a
	MF	7.3%	0.14	-0.14	0.42	0.00^a

^aP < 0.05 (between-scanner differences are significant). The Bland-Altman plots (not shown) showed no relations between mean and difference of per-subject measures of regional cartilage thickness. cLF, central lateral femur; cMF, central medial femur; LF, lateral femur; LT, lateral tibia; MF, medial femur; MT, medial tibia; RMS COV, root-mean-square coefficient of variation.

ISSN: 1478-6362