Single time point quantitation of cerebral glucose metabolism by FDG-PET without arterial sampling

EJNMMI Research

Table 1 Matrix of correlations between various FDG pharmacokinetic parameters

	A	λ1	B	λ2	C	λ3	AUC_{(0–67 min)}	Ca_{(52–67 min)}	theta_{(52–67 min)}	Age
A	1	+ 0.514 + 0.684	+ 0.555*** + 0.406	+ 0.448** + 0.147	+ 0.006 + 0.320	− 0.115 − 0.012	+ 0.180 + 0.463*	+ 0.141 + 0.508*	− 0.075 + 0.007	+ 0.284* + 0.373
λ1		1	+ 0.284 + 0.573**	+ 0.419** + 0.237	+ 0.156 + 0.192	+ 0.096 − 0.043	+ 0.019 + 0.322	− 0.148 + 0.375	+ 0.159 + 0.057	+ 0.130 + 0.508*
B			1	+ 0.503*** + 0.577*	− 0.131 + 0.159	− 0.376* – 0.096	+ 0.054 + 0.439	+ 0.241 + 0.420	− 0.187 + 0.109	+ 0.329* + 0.305
λ2				1	− 0.152 − 0.130	− 0.022 + 0.005	+ 0.032 − 0.024	− 0.157 − 0.157	− 0.127 + 0.222	+ 0.176 + 0.290
C					1	+ 0.514* + 0.737*	+ 0.510* + 0.874#**	+ 0.552*** + 0.638*	+ 0.156 + 0.548*	− 0.180 + 0.266
λ3						1	− 0.171 + 0.379	− 0.436** − 0.033	+ 0.798# + 0.846#	0.514*** + 0.084
AUC_0–67 min)							1	+ 0.678# + 0.885#	− 0.270 + 0.325	+ 0.145 + 0.324
Ca_{(52–67 min)}								1	− 0.564*** − 0.122	+ 0.244 + 0.315
theta_{(52–67 min)}									1	− 0.510*** + 0.168
Age										1

A, B, and C (arbitrary plasma concentration units) and λ1, λ2 and λ3 (fractional rate constants; min⁻¹) refer to the fitting of the tri-exponential function describing the declining blood FDG concentrations measured during 67 min. AUC_{(0–67 min)} refers to the integration of the measured blood curve to the end of the PET recordings (in arbitrary concentration units). Ca_{(52–67 min)} is the mean blood concentration of FDG measured during the final three frames of the PET recording. Theta_{(52–67 min)} is their ratio (AUC_{(0–67 min)}/Ca_{(52–67 min)}), also known as the normalized arterial input, prevailing during the final three frames of the PET recording. Age is the patient age in years. We calculated the correlations coefficients separately for groups of 52 subjects (Aarhus; upper row) and 24 subjects (Bern; lower row), where bold font indicates uncorrected significant correlation coefficients: *P < 0.05, **P < 0.002. Only ***P < 0.0002 #P < 0.00001 would survive correction for (n = 45 or 24) correlations