68 Chapter 3 by comparing DVR derived from the optimal tracer kinetic plasma input model with BPND derived from SRTM[13] . The outcome of the SRTM BPND results were subdivided into four conditions to assess the effects of flow changes in different binding levels. The four conditions were: SCD condition (almost no binding) and high, medium or low binding conditions for AD patients. Supplementary Table 1 presents the associated parametric values for each condition used for the simulations. These values were used as reference for evaluation of the simulations and are referred as the “true” value. For each level of binding, R1 values were varied to mimic 5 to 25% increase or decrease in flow (in steps of 5). Efflux rate of the tracer to the blood compartment (k2) values were also changed accordingly. In total, there were 11 different flow conditions for each level of binding (including the reference flow condition). An average TAC of the cerebellum gray was obtained from the clinical data and was used as the reference region. For each of the flow and binding conditions, 50 target tissue TACs were generated at noise level with a coefficient of variation (COV) of 5%. Subsequently, NLR based SRTM was fitted to each of these TACs to estimate the micro/macro-parameters such as R1 and BPND. In addition, SUVr80-100min was obtained for each flow condition and binding conditions. Furthermore, the average of the 50 TACs per condition were used to assess the effect of change in R1 on DVR and SUVr. Simulations were performed with noise (5%) and with almost no noise (0.05%) to the TACs, using the noise model described by Yaqub et al.[14] for both the simulations. Note that 5% COV corresponded visually to noise observed in regional TACs obtained from clinical data. SUVr values at time intervals other than 80-100 min were also obtained for each simulated flow and binding condition to evaluate the effect of the choice of scanning interval. Percentage bias for estimated parametric values (DVR and SUVr) was calculated with respect to true DVR values illustrated in Supplementary Table 1. Furthermore, percentage change for each SUVr interval was calculated with respect to true DVR values and the effect of flow was also evaluated on these measures. Supplementary Table 1. ”True” values (for each parameter) based on clinical data determined for each condition for the simulations. All micro/macro-parameters were estimated using simplified reference tissue model. SCD = subjective cognitive decline. AD = Alzheimer’s disease. BPND = binding potential. BPND k2 R1 SCD: almost no binding 0.05 0.05 0.91 AD: low binding 0.11 0.04 0.89 AD: medium binding 0.21 0.03 0.85 AD: high binding 0.32 0.08 0.88
RkJQdWJsaXNoZXIy MjY0ODMw