MRI Coil Fault Detection Using Pre-Scan Reference Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MRI systems face challenges in efficiently and accurately detecting coil faults, which can adversely affect MR image quality, without requiring extensive user intervention or causing cross-user variations.
Innovation Solution
A system utilizing a trained machine learning model to analyze reference signals collected during a pre-scan to determine coil failures, identifying the position and type of faults, and enabling real-time, automated fault detection without image acquisition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional coil fault detection methods are used, then detection can be performed, but detection time is long and user workload is high
Solution Approach 1:
The system performs fault detection using reference signals obtained during the pre-scan phase, before the actual MRI scan begins. This preliminary detection action allows faults to be identified in advance, avoiding delays during the main scanning process and reducing overall detection time.
Solution Approach 2:
The patent replaces manual inspection methods with an automated machine learning model that processes reference signals. This substitution of mechanical/manual operations with automated computational analysis significantly improves detection efficiency and reduces user workload.
2Ease of operation
If automated fault detection is implemented, then user workload is reduced, but system complexity increases
Solution Approach 1:
The system enables self-service fault detection by automatically processing reference signals through the machine learning model without requiring user intervention. The model independently analyzes the signals and provides fault determinations, reducing operational complexity from the user perspective while maintaining automated functionality.
Solution Approach 2:
The machine learning model acts as an intermediary between the raw reference signals and the fault determination. This intermediary component handles the complex analysis automatically, shielding users from system complexity while enabling sophisticated automated detection capabilities.
3Loss of time
If fault detection is performed without image acquisition, then detection time is reduced, but detection accuracy may be affected
Solution Approach 1:
The system extracts fault detection information from reference signals obtained during pre-scan, separating the detection process from the main image acquisition. This extraction allows detection to occur using dedicated reference data without the time overhead of full imaging sequences, while the machine learning model ensures accurate analysis of the extracted signals.
Data Source
Figure 1~2
Figure 3~4A
Figure 4B
AI summary
A method for coil fault detection in MRI. The method includes obtaining one or more sets of reference signals collected by a coil of a magnetic resonance imaging device in a pre-scan; obtaining a first fault detection model; determining whether the coil has a failure based on the one or more sets of reference signals and the first fault detection model.