Abstract
Background
We developed an original computer program that simulates upper limb reconstruction surgery using three‐dimensional (3D) bone models constructed from computed tomography (CT) data. However, the accuracy of a bone model has not been clarified, and radiation exposure from CT scanning posed a concern. The purpose of this study was to investigate the appropriate CT parameters required to reduce radiation exposure while maintaining the accuracy of 3D models of the forearm bones.
Methods
Twelve dry forearm bones were used to investigate the accuracy of 3D bone models created from two different CT parameters. The accuracy was evaluated by measuring: (a) the discrepancy between the models constructed from low‐ and normal‐dose CT parameters; (b) the error between actual surface data derived by a contact‐type coordinate measuring machine and a 3D bone model; and (c) the difference between a 3D bone model constructed from a bare dry bone and a model constructed from the same bone embedded in a radio‐opaque soft tissue substitute. CT dose index (CTDI) and dose‐length product (DLP) were recorded to evaluate the radiation exposure.
Results
The mean error between bone models constructed from two different CT parameters was 0.04 mm. CTDI and DLP for the normal‐radiation dose condition were 9.3 and 563 mGy/cm and those for the low‐radiation dose condition were 0.3 and 18 mGy/cm, respectively. The mean error between the bone models and scanning data from contact measurement was 0.45 mm. The mean error between a 3D model constructed from a dry bone and that constructed from the same bone embedded in a radio‐opaque soft tissue substitute was 0.06 mm.
Conclusions
3D bone models constructed from low‐radiation dose CT data demonstrated the same level of accuracy as those constructed from normal‐radiation dose data. The present simulation system can produce 3D bone models with one‐thirtieth of the normal radiation dose in the forearm. Copyright © 2009 John Wiley & Sons, Ltd.