ABSTRACT: Tumor excision is the primary treatment of aggressive or recurrent benign bone tumors and malignant bone sarcomas. This requires a surgical resection with the potential for large residual osseous defects that could be reconstructed using fresh frozen allografts. Virtual bone banks enable the creation of databases allowing a 3D pre-surgery evaluation of such allgorafts, based on segmentation of DICOM-CT images. This study demonstrates the usefulness of patient specific 3D models for an accurate host-donor allograft match. We describe one way to select the best match according to size and shape. The results suggest that a robust and reliable technique has been established. Since it is difficult to plan an allograft on a distal femur deformed by the tumor, we propose to plan the surgery on the contralateral side. Our results support this limb symmetry hypothesis. The use of this measurement protocol enables accurate selection of allografts from a contralateral healthy femur 3D CT model achieving the best match possible considering the geometry of available allograft candidate femur specimens.
Studies in health technology and informatics 01/2010; 160(Pt 2):1287-90.