A Comparative Study on Two Image Analysis Methods for the Estimation of Bone Scan Sacroiliac/Sacral Ratio

Objectives: Differences in image processing protocols may lead to variability in quantitative sacroiliac/sacral (SI/S) ratio measurements in bone scintigraphy. This study aims to compare differences in SI/S ratio values using two image analysis methods (matrix method and line method) on bone scan images. We also studied the interobserver and intraobserver variabilities between experienced and novice observers. Material and Methods: A total of 95 participants (63 with immune rheumatic diseases and 32 healthy controls) underwent Tc-99m methylene diphosphonate bone scintigraphy. SI/S ratios were calculated using two image analysis approaches: (a) the matrix method, which uses rectangular regions of interest in the sacroiliac joints (SIJs) and sacrum, and (b) the line method, which applies a single-pixel horizontal line segment across the SIJs and sacrum. Two senior observers and one junior observer performed the analyses, and interobserver and intraobserver variability were assessed using intraclass correlation coefficients (ICCs). Results: The line method consistently produced higher SI/S ratios than the matrix method across overall, right, and left measurements (all P < 0.05). Similar trends were observed in both patients and healthy subjects. Both interobserver and intraobserver agreements were high (ICC > 0.8), although minor differences were noted in less experienced readers. Conclusion: The line method yielded consistently higher SI/S ratios, indicating greater sensitivity for detecting joint activity. When transitioning between software platforms or reevaluating normal reference values, methodological differences must be acknowledged, and appropriate calibration or database revision should be considered.