MRI Guided Motion Correction for Simultaneous PET Reconstruction

This direction is merged into another direction: integrated motion correction. Its last version stays readable and citable here.

Version 1 of 120 papers27 conceptsfrozen 2026-10-04

Positron Emission Tomography (PET) is inherently a slow imaging modality. While it provides exquisite sensitivity to molecular processes, a single scan can take several minutes to acquire. During this time, patients breathe, their hearts beat, and their heads inevitably shift. This motion acts like a long-exposure photograph taken without a tripod: the resulting image is blurred, and the quantitative accuracy of the tracer concentration is compromised.

The emergence of integrated PET/MRI scanners has provided a unique solution to this problem. By using the high-speed, high-resolution structural data from MRI as a real-time "navigator," researchers can now correct for PET motion at the raw data level.

The Problem of the Moving Target

In traditional PET imaging, motion correction is often performed post-reconstruction by aligning finished images. However, Ciprian Catana and colleagues in MRI-Assisted PET Motion Correction for Neurologic Studies in an Integrated MR-PET Scanner argue that this is insufficient. When a patient moves, anatomical regions can shift into the "blind spots" or gaps between detector blocks. Post-reconstruction alignment cannot recover data lost to these physical gaps.

To solve this, researchers utilize List-Mode Rebinning. Instead of waiting for the scan to finish, the system records every single photon detection (an "event") along with a timestamp. By using simultaneous MRI to track the subject's exact position at that millisecond, the system mathematically shifts the coordinates of the raw Line-of-Response (LOR) data before reconstruction even begins. This ensures that each detected photon pair is mapped to its correct anatomical location, effectively "freezing" the motion.

MRI as a Continuous Navigator

The synergy of PET/MRI lies in temporal synchronization. As noted by Thomas A. Hope in a summary of the first ISMRM–SNMMI Workshop, MRI can serve as a continuous navigator for physiological motion.

For complex movements like cardiac contraction or respiration, simple rigid shifts are not enough. Se Young Chun and Jinsong Ouyang developed methods for MRI-Based Nonrigid Motion Correction that use "tagged" MRI to track how tissue deforms. These motion fields are integrated directly into the iterative reconstruction process—specifically the Maximum Likelihood Expectation Maximization (MLEM) framework—to warp coincidence events back to a common reference frame. This allows the system to use 100% of the acquired data, rather than "gating" (throwing away data from certain motion phases), which significantly boosts the signal-to-noise ratio.

Synchronizing the Attenuation Map

A critical but often overlooked aspect of PET imaging is Attenuation Correction (AC). To calculate the true concentration of a tracer, the system must account for how many photons were absorbed by the patient's own body. This requires an "attenuation map" of the patient's anatomy.

In a moving body, a static attenuation map becomes a liability. If the heart moves but the map stays still, the reconstruction algorithm will apply the wrong correction factors, creating artifacts that look like disease (e.g., false myocardial defects). Rong Guo and Yoann Petibon demonstrated in MR-based motion correction for cardiac PET parametric imaging that these maps must be dynamically adjusted.

Camila Muñoz further refined this in a unified framework for cardiac PET/MRI, showing that registering AC maps to the specific respiratory phase of the heart can increase image contrast by over 140%. This ensures the physical model of photon absorption remains accurate throughout the entire respiratory cycle.

Applications Across Modalities

While the bulk of this research focuses on PET/MRI, the intuition—using a structural modality to stabilize a functional one—is spreading. For instance, Thomas Kirchner and Lena Maier-hein applied a similar logic to photoacoustic imaging, interleaving ultrasound frames to calculate motion vectors that stabilize functional blood oxygenation estimates.

In clinical practice, these advances are transforming high-stakes areas:

  • Neuroimaging: Marcus Görge Ullisch showed that MRI-guided tracking allows for the resolution of tiny brain structures previously lost to blur, which is vital for early Alzheimer's detection.
  • Pediatrics: As reviewed by Ruike Chen, prospective motion correction—where the scanner's coordinate system actually follows the child's head in real-time—is making "silent" and motion-resilient imaging possible for developing brains.
  • Cardiology: Amr Abdelbaky and Ahmed Tawakol highlight how these techniques allow for the imaging of "vulnerable" plaques in tiny, rapidly moving coronary vessels.

By treating motion not as noise to be filtered out, but as a measurable physical transformation to be mathematically reversed, this research direction has turned the "blurred" limitation of PET into a high-precision quantitative tool.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_mri_guided_motion_correction_for_simulta_d8c9eceb,
  title = {MRI Guided Motion Correction for Simultaneous PET Reconstruction},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/mri-guided-motion-correction-for-simultaneous-pet-reconstruction/rdr_d8c9eceb41504832ac2d3275dc1c6b7d},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_mri_guided_motion_correction_for_simulta_d8c9eceb_v1,
  title = {MRI Guided Motion Correction for Simultaneous PET Reconstruction},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/mri-guided-motion-correction-for-simultaneous-pet-reconstruction/rdr_d8c9eceb41504832ac2d3275dc1c6b7d/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}