Breathing Compliance Classification for Adaptive Medical Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging protocols face challenges in monitoring and controlling patient compliance with desired breathing and/or breath-holding patterns, leading to motion artifacts and inefficient use of imaging devices due to delayed detection of non-compliance.
Innovation Solution
A computer-implemented method using a trained machine learning model to automatically classify patient compliance by mapping breathing information into a reduced-dimensional parameter space, allowing real-time adjustment of imaging sequences and providing non-compliance feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual evaluation of breathing information is used to determine patient compliance, then compliance assessment can be performed, but it is time-consuming and only allows determination after the imaging procedure is finished
Solution Approach 1:
The patent replaces manual mechanical evaluation of breathing information with an automated computer-implemented classification system. The system automatically receives breathing information, selects a compliance class based on mapped breathing parameters, and controls the imaging procedure accordingly, eliminating the time-consuming manual assessment process while maintaining or improving assessment accuracy.
2Reliability
If compliance monitoring is performed manually after imaging, then compliance can be evaluated, but motion artifacts and image quality issues are only detected too late requiring repeat procedures
Solution Approach 1:
The patent performs compliance classification before the imaging sequence is executed or during its execution. By determining the compliance class in advance based on mapped breathing parameters, the system can prevent motion artifacts and image quality issues before they occur, or immediately abort and repeat the sequence while the patient is still present, avoiding the need for later repeat procedures.
Solution Approach 2:
The system implements real-time feedback by continuously monitoring breathing information, mapping it to parameters, selecting compliance classes, and using this information to control the imaging procedure dynamically. This closed-loop feedback ensures image quality assurance while minimizing delays.
3Productivity
If real-time compliance monitoring is implemented, then patient compliance can be ensured during imaging, but system complexity increases with automated classification requirements
Solution Approach 1:
The patent extracts the complex compliance assessment function into a separate automated classification system that receives breathing information, maps it to parameters, and selects compliance classes independently from the main imaging control. This modular extraction manages system complexity by isolating the classification logic while maintaining high productivity through automated real-time monitoring.
4Extent of automation
If breathing information is mapped to reduced-dimensional parameter space, then compliance classification becomes computationally efficient, but information loss may occur in dimensionality reduction
Solution Approach 1:
The patent transforms raw breathing information into a reduced-dimensional parameter space by mapping breathing signals to a selected number of breathing parameters. This parameter transformation enables efficient automated compliance classification while preserving the essential characteristics needed for accurate compliance assessment, balancing computational efficiency with information retention.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Computer-implemented method for monitoring and/or controlling a medical imaging procedure on a patient (22), comprising the steps of: - receiving a breathing information (24) concerning a breathing pattern of the patient (22), - selecting a compliance class (25) from at least two possible compliance classes based on the breathing information (24), wherein at least one of the possible compliance classes corresponds to a compliance of the acquired breathing information (24) with a given desired breathing and/or breath-holding pattern (26), and - on the one hand controlling the medical imaging procedure depending on the selected compliance class (25) and/or - on the other hand outputting a non-compliance information (27) to a user and/or storing a non-compliance information (27) with an acquired medical image data (29), when the selected compliance class (25) does not indicate a compliance of the acquired breathing information (24) with the given desired breathing and/or breath-holding pattern (26).