Accurate and Robust Reconstruction of a Surface Model of the Proximal Femur From Sparse-Point Data and a Dense-Point Distribution Model for Surgical Navigation

摘要: Constructing a 3D surface model from sparse-point data is nontrivial task. Here, we report an accurate and robust approach for reconstructing of the proximal femur dense-point distribution (DPDM). The problem formulated as three-stage optimal estimation process. first stage, affine registration, to iteratively estimate scale rigid transformation between mean DPDM sparse input points. results stage are used establish point correspondences second statistical instantiation, which stably instantiates using approach. This then fed third kernel-based deformation, further refines model. Handling outliers achieved by consistently employing least trimmed squares (LTS) with roughly estimated outlier rate in all three stages. If value preferred, propose hypothesis testing procedure automatically it. We present here our validations four experiments, include 1 leave-one-out experiment, 2 experiment on evaluating handling pathology, 3 outliers, 4 models seven dry cadaver femurs clinically relevant without noise added. Our validation demonstrate performance noise. An average 95-percentile error 1.7-2.3 mm was found when reconstruct