Kalman Filtering for MRI EMI Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnet resonance imaging (MRI) systems face challenges in eliminating electromagnetic interference (EMI), which leads to signal loss, image artifacts, and calibration issues.
Innovation Solution
The use of a Kalman-like filtering process to mitigate EMI in MRI systems by utilizing both imaging and reference coils to capture image and noise data, and modeling the time-domain impulse response with a Kalman filter or smoother.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional noise removal processes are used, then some noise reduction is achieved, but performance against broadband and narrowband EMI sources is insufficient
Solution Approach 1:
The patent changes the fundamental parameters of the noise removal approach by transitioning from spectral-based methods to time-domain Kalman filtering. This involves modeling the impulse response as a state-space system with specific transition and measurement matrices, fundamentally altering how noise is characterized and removed to achieve superior performance against both broadband and narrowband EMI.
Solution Approach 2:
The Kalman filter implementation incorporates feedback mechanisms where the estimated noise is continuously refined based on the difference between predicted and actual measurements. The filter uses feedback from both imaging and reference coils to adaptively update the noise estimate, improving measurement precision while maintaining reliability in EMI removal.
2Reliability
If EMI is removed using filtering processes, then noise is reduced, but properly captured MR images or signal data may be affected
Solution Approach 1:
The patent segments the signal processing into distinct components by separating imaging signals from reference coil signals. The reference coil captures purely EMI components while the imaging coil captures both MR signals and EMI. This segmentation allows the Kalman filter to selectively remove noise from the imaging signal using the reference signal as a template, preserving MR data integrity.
Solution Approach 2:
The reference coil acts as an intermediary that captures EMI characteristics without containing MR signal information. This intermediary measurement allows the system to model and subtract EMI components from the imaging signal separately, removing noise while preserving the integrity of the MR signal data through the intermediary reference measurements.
Data Source
AI summary
Systems and methods for removing electromagnetic interference (EMI) from magnetic resonance (MR) images are disclosed. The techniques described herein include receiving first signal data for an MR scan captured using a first coil of an MR system and second signal data captured using a second coil of the MR system. The techniques include generating a filter based at least on the first signal data and the second signal data. The techniques include generating filtered signal data using the filter, the first signal data, and the second signal data.