OBJECTIVE: To estimate the dynamic rates of arterial delivery and renal drainage in renograms with the Homsy sign, using optimized computed compartmental modelling. METHODS: Eleven F-15 renograms and one F+15 renogram (using 99mTc-mercaptoacetyltriglycine) with the Homsy sign were studied, with 26 controls. Compartmental models were constructed for each renogram using components (blood, renal, bladder) linked by variables (arterial rate, drainage rate). Each model was optimized using the Marquadt least-squares method to numerical data from the time-activity curves (TACs). The model renal component at sample points along the TAC was minimized and the corresponding drainage rates calculated. RESULTS: The models were optimized, with a mean (range) r2 of 0.811 (0.68-0.88). There were continuous low drainage rates for all type 4 renograms, with no quadrupling of the flow rate, as predicted by the standard model. CONCLUSION: The standard model is unlikely to be correct and the Homsy sign is probably an artefact seen in patients with impeded urinary drainage. Computer modelling of individual renograms is feasible and can provide useful insights into the pathophysiology of renal uptake and drainage.
OBJECTIVE: To estimate the dynamic rates of arterial delivery and renal drainage in renograms with the Homsy sign, using optimized computed compartmental modelling. METHODS: Eleven F-15 renograms and one F+15 renogram (using 99mTc-mercaptoacetyltriglycine) with the Homsy sign were studied, with 26 controls. Compartmental models were constructed for each renogram using components (blood, renal, bladder) linked by variables (arterial rate, drainage rate). Each model was optimized using the Marquadt least-squares method to numerical data from the time-activity curves (TACs). The model renal component at sample points along the TAC was minimized and the corresponding drainage rates calculated. RESULTS: The models were optimized, with a mean (range) r2 of 0.811 (0.68-0.88). There were continuous low drainage rates for all type 4 renograms, with no quadrupling of the flow rate, as predicted by the standard model. CONCLUSION: The standard model is unlikely to be correct and the Homsy sign is probably an artefact seen in patients with impeded urinary drainage. Computer modelling of individual renograms is feasible and can provide useful insights into the pathophysiology of renal uptake and drainage.