Related Experiment Video
Updated: May 19, 2026
02:09
Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function
Published on: April 12, 2024
Optimization of PET-MR registrations for nonhuman primates using mutual information measures: a Multi-Transform
Christine M Sandiego1, David Weinzimmer, Richard E Carson
1Department of Biomedical Engineering, Yale University, New Haven, CT 06520, USA. christine.sandiego@yale.edu
Neuroimage
|August 29, 2012
Summary
The Multi-Transform Method (MTM) significantly improves registration success between PET and MR brain images in nonhuman primates. This new method increases successful PET-MR registrations from 77.5% to 96.7% for kinetic analysis.
Area of Science:
- Neuroimaging
- Medical Physics
- Radiochemistry
Background:
- Accurate registration of Positron Emission Tomography (PET) to Magnetic Resonance Imaging (MRI) is crucial for quantitative brain kinetic analysis in nonhuman primates.
- Traditional PET-MR registration methods using early-phase PET images (0-10 min) have a failure rate of approximately 25% due to kinetic and distribution variations.
- These registration failures can hinder neuroreceptor studies, particularly in blocking studies or with diverse radiotracers.
Purpose of the Study:
- To develop and evaluate the Multi-Transform Method (MTM) for enhancing the success rate of PET-MR image registration in nonhuman primate neuroreceptor studies.
- To compare the efficacy of two MTM algorithms (MTM-I and MTM-II) against conventional registration techniques.
- To validate the MTM approach across various PET tracers, scanner types, and primate species.
Main Methods:
- The Multi-Transform Method (MTM) was developed, involving the creation of multiple PET-MR transformations using dynamic PET data from different time intervals.
- MTM-I registered dynamic PET frames to a single MR reference, while MTM-II registered frames to multiple pre-aligned references (MR and PET).
- Normalized mutual information was used for similarity computation to select the optimal transformation, which was then assessed using visual rating scores.
Main Results:
- Successful PET-MR registrations improved from 77.5% with the conventional 0-10 min method to 85.8% with MTM-I and 96.7% with MTM-II.
- The MTM algorithms demonstrated robust performance across 120 PET datasets, 11 tracers, 3 scanner types (HR+, HRRT, Focus-220), and in baboons and rhesus monkeys.
- Visual rating scores confirmed superior spatial alignment quality with the MTM approaches compared to the standard method.
Conclusions:
- The Multi-Transform Method (MTM) offers a significant improvement in PET-MR registration success rates for nonhuman primate neuroreceptor studies.
- MTM-II, utilizing multiple reference images, achieved the highest registration success rate, proving highly effective for diverse PET imaging conditions.
- This robust technique is essential for accurate PET brain kinetic analysis, enabling more reliable neuroreceptor quantification.
