Related Experiment Video
Updated: Sep 4, 2025
05:05
Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
Published on: November 23, 2019
8.1K
"Image to patient" equal-resolution surface registration supported by a surface scanner: analysis of algorithm
Ewelina Świątek-Najwer1, Marcin Majak2, Michał Popek2
1Department of Mechanics, Materials and Biomedical Engineering, Faculty of Mechanical Engineering, Wrocław University of Science and Technology, Wrocław, Poland. ewelina.swiatek-najwer@pwr.edu.pl.
Summary
A new equal-resolution surface registration (ERSR) algorithm improves surgical instrument tracking accuracy. ERSR demonstrates statistically significant higher efficiency than landmark registration, even with lower resolution DICOM datasets.
Area of Science:
- Medical Imaging
- Computer-Aided Surgery
- Surgical Navigation
Background:
- Accurate image-to-patient registration is vital for computer-aided surgery and surgical instrument tracking.
- Existing registration methods can be sensitive to image resolution, impacting accuracy.
Purpose of the Study:
- To develop and evaluate an equal-resolution surface registration (ERSR) algorithm.
- To compare the efficiency of ERSR against standard paired-points landmark registration.
Main Methods:
- ERSR projects uniform point clouds onto DICOM and surface scanner datasets to create equal-resolution surfaces.
- Algorithm performance was tested on a phantom with titanium microscrews.
- Analysis included Target Registration Error (TRE), distance maps, and histogram evaluation, assessing DICOM resolution influence.
Main Results:
- ERSR achieved mean TREs of 0.8 ± 0.3 mm to 1.0 ± 0.7 mm across resolutions, outperforming landmark registration by at least 0.4 mm.
- Distance map histograms showed ERSR had a ~3x higher frequency in the first bin compared to landmark registration.
- Statistical analysis confirmed significant differences (Wilcoxon signed-rank test, alpha=0.05).
Conclusions:
- ERSR offers statistically significant improvements in registration efficiency compared to landmark transform registration.
- The ERSR algorithm's performance is not degraded by lower resolution CT DICOM datasets.
- ERSR exhibits greater robustness to decreased image resolution than paired-points landmark registration.
