Related Experiment Video
Updated: Oct 26, 2025
09:11
Use of MRI-ultrasound Fusion to Achieve Targeted Prostate Biopsy
Published on: April 9, 2019
21.8K
MRI-TRUS registration methodology for TRUS-guided HDR prostate brachytherapy
Philip McGeachy1,2,3, Elizabeth Watt1,2, Siraj Husain2,4
1Department of Medical Physics, Tom Baker Cancer Centre, Calgary, AB, Canada.
Journal of Applied Clinical Medical Physics
|July 28, 2021
Summary
Contour-based deformable registration in MIM software provides accurate MRI-TRUS fusion for high-dose-rate prostate brachytherapy. The Predictive Fusion workflow did not significantly enhance registration accuracy for this prostate cancer treatment.
Area of Science:
- Medical physics
- Radiation oncology
- Medical imaging
Background:
- High-dose-rate (HDR) prostate brachytherapy is a standard whole-gland treatment.
- Magnetic resonance imaging (MRI) offers superior soft-tissue contrast for improved accuracy in transrectal ultrasound (TRUS)-guided HDR prostate brachytherapy.
- The MIM treatment planning system introduced new image registration algorithms, including a Predictive Fusion workflow, to address patient positioning discrepancies between imaging modalities.
Purpose of the Study:
- To evaluate the accuracy of MIM software's image registration algorithms for MRI-TRUS fusion in HDR prostate brachytherapy.
- To assess the effectiveness of the Predictive Fusion workflow in improving MRI-TRUS registration accuracy.
Main Methods:
- Four registration methods were compared: rigid registration (RR), contour-based (CB) deformable registration, Predictive Fusion followed by RR (pfRR), and Predictive Fusion followed by CB deformable registration (pfCB).
- Prostate gland contours were delineated on both TRUS and MRI by a radiation oncologist.
- Registration accuracy was quantified using mean distance to agreement and Dice similarity coefficient.
Main Results:
- Contour-based deformable registration (CB) and Predictive Fusion followed by CB deformable registration (pfCB) demonstrated superior accuracy.
- Mean distance to agreement was lowest for CB (0.60 ± 0.08 mm) and pfCB (0.59 ± 0.06 mm).
- Dice similarity coefficients were highest for CB (0.93 ± 0.02) and pfCB (0.93 ± 0.01), indicating excellent overlap.
Conclusions:
- The contour-based deformable registration algorithm in the MIM system achieves high accuracy for MRI-TRUS fusion in HDR prostate brachytherapy.
- The Predictive Fusion workflow did not significantly improve registration accuracy compared to standard CB deformable registration.
- MIM software provides a clinically accessible platform with potential for future applications in focal therapy for prostate cancer.
