For quantification of planar and SPECT myocardial perfusion imaging lower limits of normal (LLN) are generated for each specific radiotracer from normal subjects. It is unclear how LLN for TI-201 (TI), Tc-99m-Sestamibi (MB), Tc-99m-Tetrofosmin (TF) and Tc-99m-Furifosmin (FF) compare quantitatively Accordingly we performed quantitative comparison of exercise planar (P) and SPECT (SP) in normal subjects (< 3% likelihood CAD) with TI (P:26, SP:26); MB (P:20, SP:28); TF (P:27, SP:42) and FF (P:12, SP:13). LLN (mean – 2sd) were generated using 36-points circumferential count profiles of P views and SP short axis slices, Comparisons were made by unpaired t-tests of each matched data point of 2 radiotracers, Summed t-values > 2 (p < 0,051 represent a measure for point-to-point significant difference between LLN curves, The higher the value the greater the difference. The figure shows t-values for comparison of various tracers for P and SP: LLN of various tracers are significantly different from each other and greater for P than for SP. On P imaging differences between TI and Tc-99m agents were primarily at base and anterior wall, whereas on SP no pattern existed. Differences in P imaging may be due to marked effect of interpolative background subtraction, whereas back projection and filtering produces smaller random SP differences. Thus, LLN of various agents are markedly different for P but less for SP. Radiotracer-specific normal databases should be used for quantification of P and SP myocardial perfusion imaging.
Read full abstract