A Study on Performace of Levenberg-Marquardt and CMA-ES Optimization Methods for Atlas-based 2D/3D Reconstruction

A Study on Performace of Levenberg-Marquardt and CMA-ES Optimization Methods for Atlas-based 2D/3D Reconstruction

Stať ve sborníku v databázi WoS či Scopus

In this study, we compare the performance of our previously proposed deformable 2D/3D registration approach based on the Levenberg-Marquardt optimization with methods exploiting the CMA and CMSA evolution strategies. The aim of the investigated registration is a reconstruction of a patient-specic 3D bone model from a small set of plain 2D X-ray images, which is achieved by fitting a deformable bone atlas onto the X-ray images. The comparison is focused on the robustness and the speed of the evaluated methods. The results were obtained using a large-scale data set of synthetic X-ray images. We show our method is several times faster in comparison with the approaches based on evolution strategies, while the robustness of the reconstruction is preserved. To speed-up the reconstruction process, certain parts of the registration pipeline are accelerated using graphics hardware. The median error of our proposed method was 1.12mm and the median reconstruction time was 7.2s. The median time reached by the CMA-ES and CMSA-ES methods was 48.5s and 138.5s respectively.

In this study, we compare the performance of our previously proposed deformable 2D/3D registration approach based on the Levenberg-Marquardt optimization with methods exploiting the CMA and CMSA evolution strategies. The aim of the investigated registration is a reconstruction of a patient-specic 3D bone model from a small set of plain 2D X-ray images, which is achieved by fitting a deformable bone atlas onto the X-ray images. The comparison is focused on the robustness and the speed of the evaluated methods. The results were obtained using a large-scale data set of synthetic X-ray images. We show our method is several times faster in comparison with the approaches based on evolution strategies, while the robustness of the reconstruction is preserved. To speed-up the reconstruction process, certain parts of the registration pipeline are accelerated using graphics hardware. The median error of our proposed method was 1.12mm and the median reconstruction time was 7.2s. The median time reached by the CMA-ES and CMSA-ES methods was 48.5s and 138.5s respectively.

2D/3D Reconstruction, Levenberg-Marquardt, CMA-ES, Appearance Model

2D/3D Reconstruction, Levenberg-Marquardt, CMA-ES, Appearance Model

KLÍMA, O.; CHROMÝ, A.; ZEMČÍK, P.; KLEPÁRNÍK, P.; ŠPANĚL, M.

2018

05.10.2016

Elsevier Science

Brno / Lednice

14th IFAC Conference on Programmable Devices and Embedded Systems PDES 2016 Brno, Czech Republic, 5-7 October 2016

NEUVEDEN

2405-8963

IFAC-PapersOnLine

NEUVEDEN

25

Spojené království Velké Británie a Severního Irska

121

126

6