Related Experiment Video
Updated: Apr 5, 2026
07:13
Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
Published on: October 27, 2023
1.8K
Population based ranking of frameless CT-MRI registration methods
Gabor Opposits1, Sándor A Kis1, Lajos Trón1
1Department of Nuclear Medicine, Faculty of Medicine, University of Debrecen, Hungary.
Zeitschrift Fur Medizinische Physik
|August 18, 2015
Summary
A new method using square Hellinger distance objectively ranks computed tomography-magnetic resonance imaging (CT-MRI) registration algorithms. This tool helps select the best automated method for therapy planning, improving patient care.
Area of Science:
- Medical Imaging
- Image Registration
- Radiotherapy Planning
Background:
- Simultaneous information from multiple imaging modalities (CT, MRI) is crucial in clinical practice.
- Automated registration algorithms aid in processing large datasets for radiotherapy planning.
- Objective methods are needed to rank the performance of different CT-MRI registration algorithms.
Purpose of the Study:
- Develop a quantitative method to assess CT-MRI registration algorithm performance at a population level.
- Create a performance metric within the [0,1] interval for algorithm comparison.
Main Methods:
- Collected 1051 CT-MRI image pairs from subjects.
- Established 1051 gold-standard registered images through expert manual correction.
- Evaluated various dissimilarity measures (Pearson correlation, mutual information, Hellinger distance) for performance grading.
Main Results:
- Square Hellinger distance demonstrated suitability for grading registration algorithm performance.
- This measure provides a valuable tool for developers to rank algorithms objectively.
Conclusions:
- The developed procedure offers an objective approach to identify the top-performing CT-MRI registration algorithm.
- This method aids in selecting the best algorithm for clinical applications and therapy planning.
