Dynamic Respiration Trigger for MRI Using Neural Network Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for triggering MR data acquisition in MRI face challenges due to interference from RF pulses and gradient pulses, and the variability of respiration signals across individuals, leading to inaccurate detection of the optimal data acquisition window, resulting in missed, premature, or late triggering.
Innovation Solution
A method using neural networks, specifically Temporal Convolutional Networks (TCN), Recurrent Neural Networks (RNN), or Long-Short-Term-Memory (LSTM) networks, to detect respiration direction and predict expiration amplitude peak values in real-time, adjusting the trigger point threshold dynamically based on the respiration signal amplitude, thereby improving the accuracy of data acquisition timing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a fixed amplitude threshold is used for trigger point detection, then the method is simple to implement, but the accuracy of triggering is low due to variability in respiration signals across individuals and time
Solution Approach 1:
The patent transforms the static fixed threshold into a dynamic adaptive threshold that automatically adjusts to individual respiration patterns. The system learns the subject's unique respiration characteristics through continuous monitoring and adapts the trigger threshold accordingly, resolving the contradiction between implementation simplicity and detection accuracy.
Solution Approach 2:
The patent changes the threshold parameter from a fixed value to a dynamically adjusted value based on learned respiration patterns. By modifying the threshold parameter adaptively rather than keeping it constant, the system achieves high accuracy across different individuals and time periods while maintaining automated operation.
2Measurement precision
If real-time prediction of respiration signal evolution is performed, then the accuracy of trigger point detection is improved, but the complexity of the system increases
Solution Approach 1:
The system performs self-learning by automatically analyzing the subject's own respiration signals to build a personalized prediction model. This self-service approach eliminates the need for manual calibration or complex external intervention, achieving high prediction accuracy while keeping the system relatively simple through automated adaptation.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors the actual respiration signals and compares them with predicted values. This feedback loop allows the system to refine its predictions and adjust its model parameters automatically, improving accuracy without requiring complex manual tuning or intervention.
3Speed
If pilot tone signal reception is performed simultaneously with MR data acquisition, then real-time respiration monitoring is achieved, but the signal suffers serious interference from RF pulses or gradient pulses
Solution Approach 1:
The patent performs preliminary action by acquiring respiration signals during periods when no MR data acquisition is occurring (i.e., when there is no interference from RF or gradient pulses). This pre-acquisition of clean reference signals allows the system to build accurate prediction models without being affected by harmful interference during the actual imaging process.
Data Source
AI summary
The disclosure relates to techniques for triggering magnetic resonance data acquisition. The techniques include detecting a respiration direction of a respiration signal of an acquisition subject, predicting in real time an amplitude peak value of an expiration signal in the current respiration period according to the real-time respiration signal of the acquisition subject, multiplying the amplitude peak value by a preset coefficient, and using the product as a trigger point threshold of the current respiration period. When it is determined that an expiration stage of the current respiration period is starting, the techniques also include calculating in real time or periodically the absolute value of the difference between the amplitude of the current expiration signal and the trigger point threshold currently calculated, and if the absolute value of the difference is less than a preset difference threshold, then triggering MR data acquisition.