Quantitative Sacroiliac Scintigraphy The Effect of Method of Selection of Region of Interest

Publisher Website
Google Scholar
Abstract
Various authors have advocated quantitative methods of evaluating bone scintigrams to detect sacroiliitis, while others have not found them useful. Many explanations for this disagreement have been offered, including differences in the method of case selection, ethnicity, gender and previous drug therapy. Apparently, one of the most important impediments to consistent results is the variability of selecting sacroiliac (SI) joint and reference regions of interest (ROI). The effect of ROI selection would seem particularly important because of the normal variability of radioactivity within the reference regions that have been used (sacrum, spine, iliac wing) and the inhomogeneity of activity in the SI joints. The effect of ROI selection was investigated using 5 different methods representative of, though not necessarily identical to, those found in the literature. Each method produced unique mean indices that were different for patients with ankylosing spondylitis (AS) and controls. The method of Ayres (19) proved superior (largest mean difference, smallest variance), but none worked well as a diagnostic tool because of substantial overlap of the distributions of indices of patient and control groups. ROI selection evidently is important in determining results, and quantitative scintigraphic methods in general are not effective tools for diagnosing AS. Among the possible factors limiting success, difficulty in selecting a stable reference area seems of particular importance.