Radiographically negative stress-related bone injury

Abstract

PURPOSE: To assess the diagnostic value of MR imaging relative to two-phase bone scintigraphy in radiographically negative stress-related injuries. MATERIAL AND METHODS: Sixteen consecutive patients presenting with stress-related bone injuries and normal standard radiographs were examined with two-phase bone scans and MR imaging. MR imaging protocols varied according to the region to be examined. The MR and scintigraphic images were evaluated blindly and separately by respectively two radiologists and two nuclear medicine physicians. The standard of reference was represented by a combination of clinical and scintigraphic findings as well as clinical follow-up. RESULTS: For MR imaging, sensitivity, specificity, and positive and negative predictive values for the presence of stress-related injuries for the two readers were 69/63%, 100/80%, 100/91%, and 50/40%, respectively. Interobserver agreement was good (kappa = 0.62). For scintigraphy, all abnormal and normal findings were correctly identified. CONCLUSION: For patients with clinically suspected stress-related injuries and a low probability of other active bone diseases (such as infection or neoplasm), bone scintigraphy should be the initial imaging modality.