SSOP Three-Dimensional Reconstruction of Tibia and Fibula for Applications in Biomechanical Fracture Models

Abstract

Context:  Non-fatal injuries represent a public health issue. Among them, lower limb fractures have a large impact on the costs related to orthopedic treatments. In this work, a three-dimensional reconstruction of the tibia and fibula was performed for biomechanical applications with the purpose of defining the 3D reconstruction parameters that allow reducing patients’ radiation exposure and computational costs.  Method: For the 3D reconstruction, a computerized tomography taken from a volunteer was used, as well as two software applications specialized in DICOM image reconstruction (Mimics Research and 3DSlicer). The number of images included in the volume was modified, and the results were compared. The quality of the reconstructed volumes was verified by comparing the reference volume reconstructed with the total number of images/slices vs. the modified volumes. The MeshLab software was used for this purpose. The analyzed parameters were the distance differences between the reference and the alternative models, as well as the qualitative curvature analysis. Results: The ANOVA results for the Max (maximum distance between meshes) response shows that software and slices are significant factors. However, the software-slices interaction did not have a significant influence. As for the RMS (root mean square) distance response, software, slices, and the software-slices interaction are not significant. For the Mean distance response, slices and the software-slices interaction are not significant. Nevertheless, software significantly influences the response. These results suggest a potential way to reduce the computational cost and the patient’s radiation exposure in future biomechanical and preoperatory analyses, since the same quality can be obtained by including fewer 2D images in the reconstruction. Conclusions: The reconstructed surfaces are smoother when Mimics is used, even though the same smoothness factor was employed in both software applications during the reconstruction. When 16 slices are used (retained every 16 images from the complete original model), the distance differences increased for both bones (tibia and fibula). For the RMS, reducing the number of slices and using either one of the two applications analyzed would not show any significant differences in the reconstruction, thus allowing the potential reduction of radiation exposure of the patient. Acknowledgements: The authors are grateful to Universidad Nacional de Colombia for funding the project “Estado de esfuerzos en un elemento de osteosíntesis en la consolidación de una fractura de miembro inferior”.