Table 1.

Scientific questions raised in this study and corresponding conditions under which the model is applied.

Question	Parameter sets used	Cp	${\overline{\text{C}}}_{\text{t}}$	MTT/CTH	Corresponding figure/location in the text	Supplementary information
Q1: Does our model predict a linear relation between CBF and CMRglc and between CBF and the tracer uptake, as it has been reported in the literature?	SH	Constant and equal to 500 μmol/100 mL.	bl: ${\overline{\text{C}}}_{\text{t}}$ is equal to 129 μmol/100 mL. ${\overline{\text{C}}}_{\text{t}}$ is linearly related to MTT such that it decreases by 30% between MTTbl and MTTst	MTT and CTH in the range [MTTbl-MTTst] and [CTHbl-CTHst], respectively	Figure 2
Q2: To what extent is CMRglc affected when the plasma concentration vary from hypo- to hyperglycemic levels?	SH and SR	In the range 0–3000 μmol/100 mL	${\overline{\text{C}}}_{\text{t}}$ is related to Cp according to Equations (13) and (14).	MTT = MTTbl; CTH = CTHbl	Figure 3
Q3: To what extent MTT and CTH affect CMRglc? Comparison with CMRO2.	SH	Constant and equal to 500 μmol/100 mL.	bl: ${\overline{\text{C}}}_{\text{t}}$ is equal to 129 μmol/100 mL. ${\overline{\text{C}}}_{\text{t}}$ is linearly related to MTT such that it decreases by 30% between MTTbl and MTTst.	MTT and CTH are in the range 0–2 s	Figure 4	CMRglc is much less affected by changes in CTH than CMRO2. This can be seen from iso-contours which are much more vertical for CMRglc than for CMRO2
Q4: Based on these predictions (Q3) for CMRglc and CMRO2, what is the resulting OGI and lactate production?	SH	Constant and equal to 500 μmol/100 mL.	bl: ${\overline{\text{C}}}_{\text{t}}$ is equal to 129 μmol/100 mL (Equation 14). ${\overline{\text{C}}}_{\text{t}}$ is linearly related to MTT such that it decreases by 30% between MTTbl and MTTst.	Take discrete values corresponding to the symbols (+) and (×) on Figure 4.	Figure 5	Note that ${\overline{\text{C}}}_{\text{t}}$ during baseline is calibrated to yield an OGI equal to 5.5 at baseline
Q5: Applying our model to glucose tracer (FDG), how is the LC predicted to vary as a function of the plasma concentration and between physiological states ? Are these predictions in agreement with the literature?	SH and SR	In the range 0–3000 μmol/100 mL	bl: ${\overline{\text{C}}}_{\text{t}}$ is related to Cp according to Equations (13) and (14). st: ${\overline{\text{C}}}_{\text{t}}$ is assumed to be decreased by 30% compared its value at baseline	bl: MTT = MTTbl; CTH = CTHbl st: MTT = MTTst; CTH = CTHst	Figure 6
Q6: How much would the overestimation in CMRglc be when neglecting the changes in LC?	SH	Constant and equal to 500 μmol/100 mL.	bl: ${\overline{\text{C}}}_{\text{t}}$ is set to 132 μmol/100 mL (Equation 14) and decreases by a value between 0 and 40% during stimulation	bl: MTT = MTTbl; CTH = CTHbl st: MTT = MTTst; CTH = CTHst	Figure 7	The overestimation in CMRglc increase is calculated using (Equation 28), which relates the relative changes in CMRglc, in CMRglc, app, and LC
Q7: What prediction can we make for the metabolism and the lumped constant when we apply the model to tumor cells?	SH	Constant and equal to 500 μmol/100 mL	${\overline{\text{C}}}_{\text{t}}$ during baseline is set to 132 μmol/100 mL initially and decreases as vmax_m increases.	MTT = MTTbl; CTH = CTHbl	Discussion (see Section Applying Our Model to Disease Conditions)

S_H: parameter set when the model is applied to humans: v_{max_t} = 68 μmol/100 mL_brain/min and K_T = 50 μmol/100 mL_brain. S_R: parameter set when the model is applied to rats: v_{max_t} = 136 μmol/100 mL_brain/min and K_T = 150 μmol/100 mL_brain. bl, baseline; st, stimulation. MTT_bl = 1.4 s; MTT_st = 0.81 s; CTH_bl = 1.33 s; CTH_st = 0.52 s.