Real-Time Motion Prediction Framework for Dynamic Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current motion tracking and prediction methods in dynamic imaging, such as MRI, face challenges with latency and limited accuracy, particularly in real-time applications, which can lead to imaging artifacts and complications in interventional treatments.

Innovation Solution

A novel motion prediction framework that incorporates image-based motion tracking with an adaptive filtering technique and multi-rate data fusion, enabling accurate and real-time motion prediction by processing MR images to provide feedback to clinicians and automation systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If retrospective motion tracking methods are used, then motion information can be obtained, but real-time application is not possible due to latency

Engineering Contradiction:
Improvemotion tracking accuracyVSAvoidlatency
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary motion tracking and prediction computations during the imaging acquisition process itself, rather than after. By integrating motion tracking into the real-time imaging workflow and using predictive algorithms to estimate future motion states, the system provides motion information without retrospective delay, eliminating latency while maintaining accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements a feedback mechanism where motion tracking results are continuously updated and fed back into the imaging process in real-time. This closed-loop approach allows the system to adjust and refine motion predictions dynamically, ensuring accurate real-time motion information is available for immediate clinical decision-making without retrospective processing delays.

Inventive Principle:
Principle #23Feedback

2Speed

If prospective motion tracking is used to reduce latency, then real-time feedback is possible, but undesired latencies are introduced

Engineering Contradiction:
Improvereal-time feedback speedVSAvoidmotion prediction accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The system employs dynamic adaptive filtering techniques that continuously adjust filtering parameters based on the current motion state and imaging conditions. This dynamic approach allows the system to optimize the balance between real-time performance and accuracy by adapting the level of processing to the specific clinical situation, maintaining both speed and precision.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes processing parameters dynamically based on the imaging modality and clinical requirements. By adjusting parameters such as filtering strength, prediction horizon, and processing intensity in real-time, the system optimizes the trade-off between latency reduction and motion prediction accuracy for different clinical scenarios.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If motion tracking is implemented to reduce imaging artifacts, then image quality improves, but system complexity increases

Engineering Contradiction:
Improveimage qualityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The motion tracking system is designed to be modality-agnostic and can be integrated with multiple imaging techniques (MRI, CT, ultrasound, etc.). By creating a universal motion tracking platform that works across different imaging modalities, the system reduces overall complexity compared to implementing separate specialized tracking systems for each modality, while consistently improving image quality through artifact reduction.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3595525B1Systems and methods for real-time motion prediction in dynamic imaging
Publication Date: 2023.09.27 RGT UNIV OF CALIFORNIA
  • EP3595525B1 patent drawingFigure 1
  • EP3595525B1 patent drawingFigure 2
  • EP3595525B1 patent drawingFigure 3A

AI summary

Systems and methods for predicting motion of a target using imaging are provided. In one aspect, a method includes receiving image data, acquired using an imaging system, corresponding to a region of interest ("ROI") in a subject, and generating a set of reconstructed images from the image data. The method also includes processing the set of reconstructed images to obtain motion information associated with a target in the ROI, and applying the motion information in a motion prediction framework to estimate a predicted motion of the target. The method further includes generating a report based on the predicted motion estimated.