PubMed · 16393790
A point-selection algorithm based on spatial-stiffness analysis of rigid registration.
Abstract
OBJECTIVE: We propose a model of shape-based registration that leads to a task-specific algorithm for preoperatively selecting a set of model registration points. MATERIALS AND METHODS: We performed five sets of computer simulations using registration points generated by our algorithm and two noise amplification index (NAI) algorithms on the basis of the research of Simon 20. We used several different bone surface models (distal radius, proximal femur and tibia) computed from CT images of patient volunteers. The number of registration points used varied between 6 and 30. RESULTS: Our algorithm was faster than the NAI-based algorithms by factors of approximately 4 and 200. It had equal or better performance in terms of target registration error (TRE) when compared with the other algorithms. Our simulations also showed that point selection can have a large effect on TRE behavior; in particular, poor point selection does not necessarily decrease TRE as more registration points are added. CONCLUSIONS: Our point-selection algorithm produces model registration points with similar or better TRE behavior than the NAI-based algorithms we tested, and it does so with significantly less computation time.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
B Ma, R E Ellis. 2005. A point-selection algorithm based on spatial-stiffness analysis of rigid registration.. https://doi.org/10.3109/10929080500230320. https://pubmed.health/record/pubmed%3A16393790/a-point-selection-algorithm-based-on-spatial-stiffness-analysis-of-rigid-registration
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite PubMed Health by Indiciae: BibTeX · RIS.