Multimodality imaging for precise localization of craniofacial osteomyelitis

Abstract

Functional imaging identifies areas of abnormal bone turnover, providing a useful adjunct in the treatment of osteomyelitis and bone tumors. The low resolution and lack of anatomical detail limit the application of bone scans in craniofacial surgery, however. Multimodality image registration addresses this problem by fusing functional images (single photon emission computed tomography [SPECT]) to high-resolution structural images (computed tomography [CT]) for precise anatomical delineation of bone activity. This article describes a technique for spatial registration of CT and SPECT images to provide precise anatomical delineation of abnormal bone turnover, thereby guiding the extent of resection in the management of craniofacial osteomyelitis. Standard CT and SPECT imaging protocols were used in imaging the skull from the vertex to the mentum. Image data were imported into Analyze (Biomedical Imaging Resource; Mayo Foundation, Rochester, MN) on a dedicated Windows NT (Microsoft Corporation, Redmond, WA) workstation. Using the CT data, the craniofacial skeleton, osteotomy segments, and bone grafts were interactively mapped out. Consecutive axial slices were then reconstructed to form a three-dimensional volume of interest. The CT-derived volume of interest was registered to the technetium Tc 99m-methylene diphosphonate SPECT scan using the Analyze program to provide a fused multimodality image. The imaging technique was used to localize osteomyelitis in a complex craniofacial reconstruction. The fused images guided the extent of resection during surgery, and postoperative microbiological and histological testing confirmed the diagnosis. Multimodality image registration provides a readily available method to relate facial skeletal anatomy and physiology. This technique is valuable in planning and monitoring therapeutic interventions in clinical conditions in which bone turnover is abnormal.