SummaryThe aim of this study was to assess the influence of background subtraction (BS) on estimation of differential kidney function (DF) on the static scintigraphy with 99mTc dimercaptosuccinic acid (DMSA) and dynamic scintigraphy with 99mTc mercaptoacetyltriglicine (MAG3) and to establish possible differences between DF values estimated with these methods. Patients, methods: Patients (n = 106) were selected among those scheduled to static and dynamic scintigraphies within 3 months, with no interim clinical and laboratory changes, regardless the kidney pathology. DF was estimated according to the uptake ratio method. Four background regions of interest (ROIs) were applied, identical for both studies, and DF values were recalculated after BS. The corrected values were compared to the values before correction, separately for DMSA and MAG3, and between the studies. The results showed that ROIs used introduce variable results for the same patients, predominantly when noncorrected DF values were <45%. There were no significant differences between DF values (corrected and noncorrected) obtained from static and dynamic scintigraphy in all groups of patients. Since numerous reasons can bring to the errors in DF estimation when BS is used, the conclusions are that it would probably be more accurate to avoid BS, particularly when DF values are compared in a patient follow-up, and when kidney function is normal. BS should be used, but always in the same way, only when there is a significant difference in kidney size, or when DF is <25%, since background activity is then considerable. MAG3 and DMSA can be equally used for DF estimation and their results compared in patient follow-up.
Combining Background Subtraction and Convolutional Neural Network for Anomaly Detection in Pumping-Unit Surveillance
