J Nucl Med. 2008; 49 (Supplement 1):157P
InfoSNM Computer PresentationsInfoSNM II |
Development of a software application for quantitative processing of nuclear renography
Russell Folks1,
Ernest Garcia1 and
Andrew Taylor1
1 Radiology, Emory University School of Medicine, Atlanta, Georgia
660
Learning Objectives: 1. Name 3 indications for performing a quantitative nuclear renographic study. 2. Describe ways to improve reproducibility in renal image analysis. 3. List 4 ways in which a computer program can perform quality control monitoring in renographic analysis.
Abstract (summary): A new software application has been developed for processing nuclear renographic images, calculating clinically-relevant parameters of renal function and presenting summarized results for interpretation. The application runs on a personal computer under Microsoft Windows XP. The application can handle dynamic and static images from a baseline renal function study or a diuretic study, in interfile format. For the baseline, automatic regions of interest (ROI) are generated for the whole kidneys, renal cortices, background and prevoid and postvoid bladder. The user can manually adjust any ROI, manually draw an injection site ROI, and indicate the time of bolus arrival in the kidney. Reframed images and time-activity curves are generated, and up to 20 different renal functional parameters are calculated, including camera-based clearance. The application automatically monitors 30 different items related to image and processing quality. Processed results are presented onscreen, with normal ranges for calculated values. Options are provided for saving the results to disk as text or in graphic format. For a diuretic study, manual kidney and renal pelvis ROIs can be drawn, and up to 14 additional values related to renal obstruction are calculated. The diuretic study can be related to its baseline study. An expert system for scan interpretation (previously reported) can be accessed for decision support. Processing, review options, quality assurance functions and expert system connectivity will be demonstrated.
