Is deconvolution applicable to renography?

The feasibility of deconvolution depends on many factors, but the technique cannot provide accurate results if the maximal transit time (MaxTT) is longer than the duration of the acquisition. This study evaluated whether, on the basis of a 20 min renogram, it is possible to predict in which cases the MaxTT will exceed 20 min. Renograms of various shapes were simulated by convolution of a plasma disappearance curve and various created retention functions with a mean transit time (MTT) ranging from 3 to 23 min. The values of MaxTT were then derived from the created curves and compared to three parameters of transit measured on the renograms: the time to reach the maximum of the curve (Tmax), the output efficiency at 20 min (OE20), and the normalized residual activity at 20 min (NORA20). The proportion of retention functions (n=390) with MaxTT>20 min increased with increasing Tmax (e.g. 9% for 6< or =Tmax<10 min, and 34% for 11< or =Tmax<15 min), increasing NORA20 (e.g. 20% for 1.4< or =NORA20<3.0, and 84% for 3.0< or =NORA20<5.0) and decreasing OE20 (19% for 50% <OE20< or =75%, and 76% for 25% <OE20< or =50%). Use of Tmax, OE20 and NORA20 doesn't allow the differentiation of cases with a MaxTT longer or shorter than 20 min. Deconvolution can paradoxically only be used in cases of normal transit.