ABSTRACT
Nuclear medicine departments use the camera-based method for determining glomerular filtration rate (GFR) with 99mTc-DPTA. Camera based techniques provide a fast and convenient way of determining GFR with excellent reproducubility however its accuracy remains in question. The accuracy of camera-based renal scintigraphy depends on an attenuation correlation from estimating for renal depth and an attenuation correction from estimating for renal depth and an attenuation coefficient. Algorithms were formulated by Tonnesen, Taylor, and Inoue to calculate the estimated renal depth through multiple stepwise linear regression analysis. The goal of this study was to analyze the accuracy of these algorithms in Filipino patients. Renal depth was determined from CT scans of 41 consecutive patients. We calculated the mean absolute error of the estimated kidney depths and compared them with each other to determine the accuracy of each algorithm. The Tonnesen algorithm had the highest mean absolute error of 1.25 cm plus minus 0.88 cm in the left kidney estimates while the Taylor algorithm had the lowest mean absolute error of 0.81 cm plus minus 0.58 cm in the left kidney estimate and 1.04 cm plus minus 0.82 cm in the right kidney estimate (p-value = 0.01 left kidney estimates and p-value = 0.02 right kidney estimates). The Tonnesen algorithm was the least accurate in predicting kidney depth. There was no significant difference between the Taylor and Inoue algorithms, which were more accurate in predicting kidney depth