Detection of small periprosthetic bone defects after total knee arthroplasty.

Abstract

Large bone defect around total knee prostheses is among the most critical challenges in revision surgery. However, it is difficult to detect bone defects around a prosthesis in early stage. We compared the efficacy of the detection of small bone defects between fluoroscopically guided plain radiography, CT, MRI, and a novel tomographic technique (tomosynthesis) using the six pig knee models. No bone defects were detected with plain radiography and MRI. The sensitivity and specificity of CT were 61.5% and 64.1%, respectively. The sensitivity and specificity of tomosynthesis were 85.4% and 87.2%, respectively. The radiation dose of tomosynthesis was 6% of that of CT. The cost of tomosynthesis was 28% of that of CT. Tomosynthesis was superior in terms of diagnosis, radiation dose, and cost.