Real-Time Motion Prediction for Dynamic MRI
- Technology Benefits
- The motion information is extracted based on only real-time images, and there is no need for additional sensors, contrast markers, or hardware modification Accurate prediction results are generated to overcome system latencies The real-time processed feedback information can be used for autonomous and continuous guidance/control of interventional procedures The real-time processed feedback information can be used to update MRI scan parameters, or improve reconstruction, of MR images for non-interventional diagnostic applications
- Technology Application
- Real-time MRI MRI-guided interventions (e.g. needle targeting, non-invasive ablation) Non-interventional diagnostic dynamic MRI X-ray, CT, PET, US, optical and other forms of imaging
- Detailed Technology Description
- Researchers at UCLA have developed a novel real-time motion prediction algorithm that incorporates MR image-based motion tracking, adaptive filtering, and adaptive calibration to predict the motion state at a particular time point. This algorithm generates accurate motion information (e.g., coordinates) of the moving targets (e.g., tissues and devices) from MRI in real time, and the motion information can be used for the adjustment of MR image planes and parameters, compensation of motion effects to improve MR image quality, or it can be sent to the physicians and/or interventional systems as feedback for MRI-guided interventional procedures.
- Others
-
State Of Development
Concepts and initial prototype.
Background
Magnetic resonance imaging (MRI) has the advantage of producing improved soft tissue contrast compared to other imaging modalities to perform diagnostic decisions and guide interventions without the use of ionizing radiation. However, MR image quality and interpretation are often compromised by motion and flow artifacts, especially in regions of the body with constant physiological motion, such as the chest, abdomen, and pelvis. These motion artifacts can cause considerable signal change, interfering with image quality and image interpretation in dynamic and static MRI. Conventional MRI has low frame rate and faces challenges to provide high-quality images that avoid motion artifacts or resolve dynamic processes. In addition, there is currently no available online processing algorithm for real-time MRI to provide feedback to physicians and interventional systems during MRI-guided interventions.
Tech ID/UC Case
29476/2017-505-0
Related Cases
2017-505-0
- *Abstract
-
UCLA researchers in the Department of Bioengineering and Department of Radiological Sciences have developed a novel motion prediction algorithm using MRI-based motion tracking to provide accurate and real-time motion information for dynamic MRI and MRI-guided interventions.
- *Principal Investigator
-
Name: Xinzhou Li
Department:
Name: Holden Wu
Department:
- Country/Region
- USA
