Skip to content

Covariance Matrix Adaptation Evolution Strategy (CMA-ES) for 2D/3D registration

Notifications You must be signed in to change notification settings

m1nhengChen/CMAES-reg

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

24 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

News

  • 2024/5/17 Provides a fully differentiable optimization-based approach implemented by PyTorch’s stochastic gradient descent optimizer, which is used as a baseline in our experiments. ( Reference: Link )
  • 2024/1/3 Provides more similarity measures including multi-scale normalized cross-correlation, etc. ( Reference: DiffPose )

CMAES-reg

Using Covariance Matrix Adaptation Evolution Strategy (CMA-ES) for 2D/3D registration.The procedure is single-resolution and the DRR module is implmented by projective spatial transformers.

  • The result is evaluated on simulation data. To evaluate the registration, we follow the standardized evaluation methodology to report the Mean Target Registration Error(mTRE) as follows:
mTRE(mm) SR Reg.
time
95th 75th 50th
Initial 205.81 172.05 135.12 0.0 N/A
CMA-ES 31.72 8.37 5.39 49.4 21.2s
  • The mTRE results are reported in forms of the 50th, 75th, and 95th percentiles to demonstrate the robustness of our methods. In addition, we also report the success rate (SR) and average registration time, where SR is defined as the percentage of the tested cases with a TRE smaller than 10 mm.

Although this framework is single-resolution, it can be easily changed to multi-resolution. Feel free to replace its similarity measure and tune the hyperparameters.

Citation

If you find this work useful in your research, please cite the appropriate papers:

@article{chen2024optimization,
  title={An Optimization-based Baseline for Rigid 2D/3D Registration Applied to Spine Surgical Navigation Using CMA-ES},
  author={Chen, Minheng and Li, Tonglong and Zhang, Zhirun and Kong, Youyong},
  journal={arXiv preprint arXiv:2402.05642},
  year={2024}
}
@article{zhang2024introducing,
  title={Introducing Learning Rate Adaptation CMA-ES into Rigid 2D/3D Registration for Robotic Navigation in Spine Surgery},
  author={Zhang, Zhirun and Chen, Minheng},
  journal={arXiv preprint arXiv:2405.10186},
  year={2024}
}

About

Covariance Matrix Adaptation Evolution Strategy (CMA-ES) for 2D/3D registration

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages