Three-dimensional virtual bone bank system for selecting massive bone allograft in orthopaedic oncology

Purpose Although structural bone allografts have been used for years to treat large defects caused by tumour or trauma, selecting the most appropriate allograft is still challenging. The objectives of this study were to: (1) describe the establishment of a visual bone bank system and workflow of allograft selection, and (2) show mid-term follow-up results of patients after allograft implantation. Methods Allografts were scanned and stored in Digital Imaging and Communications in Medicine (DICOM) files. Then, image segmentation was conducted and 3D model reconstructed to establish a visual bone bank system. Based on the volume registration method, allografts were selected after a careful matching process. From November 2010 to June 2013, with the help of the Computer-assisted Orthopaedic Surgery (CAOS) navigation system, the allografts were implanted in 14 patients to fill defects after tumour resection. Results By combining the virtual bone bank and CAOS, selection time was reduced and matching accuracy was increased. After 27.5 months of follow-up, the mean Musculoskeletal Tumor Society (MSTS) 93 functional score was 25.7 ± 1.1 points. Except for two patients with pulmonary metastases, 12 patents were alive without evidence of disease at the time this report was written. Conclusions The virtual bone bank system was helpful for allograft selection, tumour excision and bone reconstruction, thereby improving the safety and effectiveness of limb-salvage surgery.