Imaging pitfalls of interbody spinal implants.

STUDY DESIGN: A variety of interbody implants were imaged by computed tomography and plain radiography within cadaveric spines to evaluate their basic imaging characteristics. OBJECTIVES: Sources of interpretation error by both computed tomography and plain radiography of interbody implants were investigated. SUMMARY OF BACKGROUND DATA: Lucencies have been reported around bone dowel implants in the postoperative period, which have been shown to resolve. The diagnosis of fusion through metallic implants has been difficult with both false-positive and false-negative results. The literature of imaging these implants is very sparse. METHODS: Four interbody constructs were placed in cadaveric spines under different conditions and imaged using both computed tomography and plain film radiography. RESULTS: Plain radiographs could not predict the presence of intraimplant bone whereas computed tomography was accurate. Metallic implants had a 1-3-mm computed tomography artifact limiting peri-implant interpretation. Lucencies could be seen on computed tomography but not on plain radiographs. However, the opposite was also seen. Lucencies around nonmetallic implants were more visible by a 4:1 ratio. CONCLUSIONS: Potential errors in interpretation were identified including assessment for bridging bone, assessment for lucency, and obscuration of peri-implant detail from metallic artifact. Lucencies were more visible with nonmetallic implants than with metallic constructs. Plain radiograph analysis of metallic implants tended to underestimate lucencies, whereas analysis of nonmetallic implants tended to overestimate lucencies.

In vivo regional cerebral blood volume: quantitative assessment with 3D T1-weighted pre- and postcontrast MR imaging.

PURPOSE: To assess the use of pre-and postcontrast three-dimensional (3D) T1-weighted images to obtain quantitative regional cerebral blood volume (rCBV) maps and to delineate the vascular system of the brain. MATERIALS AND METHODS: Sagittal 3D pre-and postcontrast T1-weighted images were acquired in 10 patients and two volunteers. The images were processed with a knowledge-based segmentation algorithm to obtain independent blood volume maps of gray matter (GM), white matter (WM), and blood vessels and to quantitate rCBV in various cortical and deep cortical structures. Localized maximum intensity projection (MIP)-rCBV images were used to reveal the brain's venous system and vascularity in and around any lesions present. RESULTS: Group-averaged (n = 8) rCBV maps from both GM (4.80% +/- 0.37) and WM (2.02% +/- 0.14) yielded a ratio of GM-to-WM rCBV of 2.38 +/- 0.20. The ratio of rCBV between cortical and deep cortical GM was 1.15. rCBV maps showed better vascular contrast than did the postcontrast images in delineation of blood vessels and in the lesions present in two patients. A localized MIP-rCBV image had more than five times higher vascular contrast-to-noise ratio than that of the postcontrast MIP image. CONCLUSION: rCBV maps can be obtained in a standard clinical setting and can be used to reveal information on local blood volume.