PET Motion Correction via MR K-space Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current PET data acquisition methods face challenges in motion correction, particularly during abdomen scans, where patient breathing leads to motion artifacts in PET images, affecting diagnosis quality.
Innovation Solution
A method involving the acquisition of MR data to determine a motion model, which provides correction specifications for PET data, using MR raw data converted into a simple value to quickly determine the current motion state, and applying this model to correct PET images by transforming them into a standardized motion state.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MR data is acquired simultaneously with PET data to determine motion state for correction, then motion correction accuracy is improved, but measurement time and system complexity increase
Solution Approach 1:
The patent extracts only the essential motion information from MR data by converting MR raw data into a simple value that represents the current motion state, rather than processing complete MR images. This selective extraction maintains motion correction accuracy while significantly reducing processing time and computational requirements.
Solution Approach 2:
The motion model is determined in advance from MR data acquired during a preliminary measurement phase. This pre-established motion model can then be applied to correct PET data without requiring real-time MR processing during the actual PET acquisition, thereby reducing measurement time while preserving correction accuracy.
2Productivity
If MR raw data is converted into a simple value to quickly determine motion state, then processing speed is improved, but information loss may occur
Solution Approach 1:
The patent extracts only the essential motion information from MR raw data by converting it into a simple value that represents the current motion state, rather than processing complete MR images. This selective extraction maintains motion correction accuracy while significantly reducing processing time and computational requirements.
Solution Approach 2:
The patent transforms complex MR raw data into a simplified parameter (simple value) that captures the essential motion state information. This parameter transformation maintains the necessary information for motion correction while enabling rapid processing and real-time application during PET acquisition.
3Manufacturing precision
If a motion model is created to correct PET data for different motion states, then image quality is improved, but device complexity increases
Solution Approach 1:
The patent transforms complex MR raw data into a simplified parameter (simple value) that captures the essential motion state information. This parameter transformation maintains the necessary information for motion correction while enabling rapid processing and real-time application during PET acquisition.
Solution Approach 2:
The motion correction system is designed to handle multiple motion states (different breathing phases, cardiac cycles) using a unified motion model framework. This multi-functional approach allows the same system to correct various types of motion artifacts without requiring separate specialized systems for each motion type.
Data Source
AI summary
A method is disclosed for creating a motion correction for PET data acquired by a PET system from a volume segment of an examination object. The method includes acquisition of MR data within the volume segment by the magnetic resonance system; and determination of a motion model of a motion within the volume segment as a function of the MR data. The motion model, as a function of a respective motion state of the motion, provides a correction specification for PET data which is acquired during this motion state. During acquisition of the MR data, specific MR data is acquired in the center of the k-space or of a straight-line segment which passes through the center of the k-space. The MR data determined is converted by a mathematical function into one value, as a function of which the respective motion state is determined.