BACKGROUND: Parametric kidney scintigraphy gives the possibility of regional function distribution assessment of these organs; the clinical application of the method has thus far been limited. The usefulness of the method for the assessment of postinflammatory scars and diabetic nephropathy has been demonstrated. Preliminary data also indicate that this type of imaging allows the assessment of kidney function after extracorporeal shock wave lithotripsy. The aim of this study was the evaluation of the reproducibility of semiquantitative parametric kidney image evaluation.
MATERIAL AND METHODS: The results of 98 dynamic kidney scintigraphic series were evaluated, obtained from examination of 44 patients (20 males, 24 females) with nephrolithiasis, who had been treated by means of lithotripsy. The semiquantitative assessment involved conventional renoscintigraphic images obtained from summation of scintigraphic serial records in the secretory phase, and parametric clearance images. A 5-level score was applied for assessment of both types of images, based upon numbers of detected defects of the regional function (0 - no defects; 1, 2, 3 - for 1, 2 or 3 defects, and 4 for higher numbers). Altogether, 196 kidney images were evaluated. The assessment was performed independently by 2 observers: A - an experienced specialist in nuclear medicine, who evaluated the images twice, and B - a resident physician with limited experience in the field.
RESULTS: The agreement between the two evaluations by the specialist (intra-observer test) reached a level of 96% for conventional images and 90% for the parametric ones. In the inter-observer test, full agreement reached the levels of 84% and 71% for conventional and parametric images, respectively. If ± 1 degree of the image score was taken as factual agreement, the intra-observer concordance reached 100%, and for inter-observer comparison, the agreement reached 99% and 97% for conventional and parametric images, respectively.
CONCLUSIONS: The concordance of image assessment for conventional and parametric images is very good. In observations by two physicians, a somewhat closer agreement was reached for conventional than for parametric images; the difference between the two series of image assessments was small.