OBJECTIVE: Reformatted CT-based image data may be of use in evaluation of new bone formation around massive bone replacement implants, but the impact of metal-induced artifacts on the accuracy of image reconstruction is unknown. This study was designed to evaluate the accuracy and precision of quantitative image reconstruction in the presence of a titanium implant. MATERIALS AND METHODS: Model porous-coated titanium implants were harvested from 23 dogs 12 weeks after surgical implantation. Contiguous 1.5 mm transverse CT scans were done first with the titanium implant in place, then repeated after replacement of the implant with an acrylic spacer. Microradiographs of sections cut at defined locations were used to establish gray level thresholds for image reconstruction and to validate the accuracy of CT-based image data. Reformatted image data were used to determine periprosthetic bone volume (mm3) and area of interfacial contact between bone and implant (mm2). RESULTS: The CT-based data derived with either implant material had an accuracy of 83-88% and a precision coefficient of variation (CV) of 2-4% for both volume and contact area variables. Presence of titanium did not seriously affect the quality of images obtained, and results were strongly correlated with those obtained with acrylic. CONCLUSION: Accurate volumetric data derived from CT-based images of periprosthetic new bone formation can be obtained in the presence of a titanium endoprosthesis. Improved demonstration of spatial relationships of bone and implant may improve postoperative evaluation.
OBJECTIVE: Reformatted CT-based image data may be of use in evaluation of new bone formation around massive bone replacement implants, but the impact of metal-induced artifacts on the accuracy of image reconstruction is unknown. This study was designed to evaluate the accuracy and precision of quantitative image reconstruction in the presence of a titanium implant. MATERIALS AND METHODS: Model porous-coated titanium implants were harvested from 23 dogs 12 weeks after surgical implantation. Contiguous 1.5 mm transverse CT scans were done first with the titanium implant in place, then repeated after replacement of the implant with an acrylic spacer. Microradiographs of sections cut at defined locations were used to establish gray level thresholds for image reconstruction and to validate the accuracy of CT-based image data. Reformatted image data were used to determine periprosthetic bone volume (mm3) and area of interfacial contact between bone and implant (mm2). RESULTS: The CT-based data derived with either implant material had an accuracy of 83-88% and a precision coefficient of variation (CV) of 2-4% for both volume and contact area variables. Presence of titanium did not seriously affect the quality of images obtained, and results were strongly correlated with those obtained with acrylic. CONCLUSION: Accurate volumetric data derived from CT-based images of periprosthetic new bone formation can be obtained in the presence of a titanium endoprosthesis. Improved demonstration of spatial relationships of bone and implant may improve postoperative evaluation.