Article Figures & Data
Figures
- FIGURE 1.
Example of curve fitting in 1 patient using monoexponential (A and C) and alternative biphasic (B and D) models applied to all (A and B) and initial (C and D) data. In each plot, only black points are used for curve fitting. Dashed lines show fitted curves. Figure shows ability of respective models to fit data and to predict course of time–activity curve when fitted to limited number of initial data points. Values of estimated parameters are t1 = 0.89 d, ln A(0) = 6.29, Tef = 2.29 d, τ = 0.22 d (A); C1 = 0.41, C2 = 5.80, C3 = −0.24, τ = 0.37 d (B); t1 = 0.89 d, ln A(0) = 5.74, Tef = 7.07 d, τ = 0.69 d (C); and C1 = 0.34, C2 = 5.72, C3 = −0.21, τ = 0.36 d (D). Average ln errors of fitting are 0.06 (A) and 0.03 (B). Average ln errors of prediction are 2.96 (C) and 0.08 (D). When only 1 measurement a day is evaluated, the errors increase to 0.07 (A), 0.06 (B), 3.88 (C), and 0.43 (D), respectively. exp. = exponential.
- FIGURE 2.
Mean values and ratios of τd and τt obtained by monoexponential (M) and biphasic (B) models for 633 routine records with infrequent sampling (M633, B633) and for 71 experimental records with frequent sampling (M71, B71).
Tables
- TABLE 1
Logarithms of Estimates of Residence Times τd After Diagnostic Administration of 131I
Model n Arithmetic mean of ln τd SD of ln τd Effect of model Effect of sampling Monoexponential 633 −3.71 1.32 Biphasic 633 −3.71 1.21 P > 0.05 Monoexponential 71 −3.02 2.23 P < 0.05 Biphasic 71 −2.88 1.68 P > 0.05 P < 0.01 Effect of model shows difference between monoexponential and biphasic models in each group of patients. Effect is evaluated by paired t test. Effect of sampling shows difference between 2 groups of patients with different sampling for monoexponential and alternative models. Effect is evaluated by standard t test for 2 arithmetic means.
- TABLE 2
Logarithms of Estimates of Residence Times τt After Therapeutic Administration of 131I
Model n Arithmetic mean of ln τt SD of ln τt Effect of model Effect of sampling Monoexponential 633 −5.43 1.08 Biphasic 633 −4.77 0.94 P < 0.01 Monoexponential 71 −3.79 1.06 P < 0.01 Biphasic 71 −3.27 1.06 P < 0.01 P < 0.01 Effect of model shows difference between monoexponential and biphasic models in each group of patients. Effect is evaluated by paired t test. Effect of sampling shows difference between 2 groups of patients with different sampling for monoexponential and alternative models. Effect is evaluated by standard t test for 2 arithmetic means.
- TABLE 3
Paired Differences Between Logarithms of Estimates of Residence Times ln τd − ln τt
Model n Arithmetic mean of ln τd − ln τt SD of ln τd − ln τt Paired t test of ln τd − ln τt Effect of model Effect of sampling Monoexponential 633 1.72 0.95 P < 0.01 Biphasic 633 1.07 0.76 P < 0.01 P < 0.01 Monoexponential 71 0.77 1.31 P < 0.01 P < 0.01 Biphasic 71 0.38 0.75 P < 0.01 P < 0.01 P < 0.01 Differences ln τd − ln τt in each row are evaluated by paired t test. Effect of model shows difference between monoexponential and biphasic models in each group of patients. Effect is evaluated by paired t test. Effect of sampling shows difference between 2 groups of patients with different sampling for monoexponential and alternative models. Effect is evaluated by standard t test for 2 arithmetic means.
Model n Measurement twice a day Measurement once a day Arith. mean of ln error SD of ln error Effect of model Arith. mean of ln error SD of ln error Effect of model Monoexponential 64 1.33 1.28 2.14 0.89 Biphasic 64 0.58 1.40 P < 0.01 2.15 0.89 P > 0.05 Arith. = arithmetic.
Effect of model shows difference between monoexponential and biphasic models evaluated by paired t test.
Model n Measurement twice a day Measurement once a day Arith. mean of ln error SD of ln error Effect of model Arith. mean of ln error SD of ln error Effect of model Monoexponential 64 2.87 0.91 1.76 0.61 Biphasic 64 1.33 0.97 P < 0.01 1.62 0.85 P < 0.05 Arith. = arithmetic.
Effect of model shows difference between monoexponential and biphasic models evaluated by paired t test.
