Deep Learning MRI Image Quality Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MRI image quality assessment methods are inefficient and lack real-time capabilities, often resulting in low-quality images being identified only after the examination is completed, and they fail to account for task-specific quality considerations, leading to inconsistent and costly human review processes.
Innovation Solution
A deep learning-based method for real-time image quality assessment using a computer system that provides MR image data to a trained model for immediate classification, allowing for corrective action during the scan and reducing the need for repeat procedures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image quality assessment methods are used, then expert review can be performed, but the process is slow and not real-time
Solution Approach 1:
The patent replaces the mechanical system of manual expert review with an automated deep learning model. The CNN-based image quality assessment system processes MRI images automatically, eliminating the need for human experts to manually evaluate each image, thereby achieving real-time assessment while maintaining high accuracy.
Solution Approach 2:
The system enables self-service by allowing the MRI scanning system to automatically assess its own image quality without external expert intervention. The deep learning model is integrated into the workflow to provide autonomous quality evaluation, reducing dependency on human experts and enabling immediate feedback.
2Reliability
If traditional two-step quality assessment process is used, then diagnostic quality can be evaluated, but the process requires expert input and is slow
Solution Approach 1:
The patent merges the two-step process of condition identification and quality assessment into a single integrated step. The deep learning model simultaneously evaluates both the diagnostic quality and task-specific adequacy of MRI images in one operation, eliminating the need for separate expert review steps and reducing overall process complexity.
Solution Approach 2:
The assessment system is designed to be universal and task-specific simultaneously. The deep learning model can evaluate multiple quality metrics including diagnostic quality, task-specific adequacy, and various image artifacts in a single assessment, making the system versatile for different MRI applications without requiring separate specialized processes.
3Quantity of substance
If MRI scanning is performed with long acquisition time and multiple breath-holds, then diagnostic data can be obtained, but image quality becomes inconsistent and robustness decreases
Solution Approach 1:
The system performs preliminary quality assessment during the scanning process itself, before the patient leaves. This real-time monitoring allows operators to detect quality issues early and take corrective actions such as adjusting scanning parameters or performing additional scans while the patient is still present, thereby improving image quality consistency without requiring excessively long acquisition times.
Solution Approach 2:
The patent implements a feedback mechanism where the deep learning model continuously evaluates image quality during and after scanning. This feedback loop provides immediate information about image quality consistency, allowing operators to adjust scanning protocols in real-time or identify patterns that require protocol modifications to improve overall reliability and reduce the need for multiple breath-holds.
Data Source
AI summary
A method for real-time assessment of image information employing a computer that includes one or more processors includes obtaining, via a scanner, a magnetic resonance (MR) image of a region-of-interest, and providing data corresponding to at least a portion of the MR image to a deep learning model, where the deep learning model is previously trained based on one or more sets of training data. The method further includes assessing the data using the deep learning model and data obtained from an image quality database to obtain an assessed image quality value, and formulating an image quality classification in response to the assessed image quality value. The method additionally includes outputting the image quality classification within a predetermined time period from an initial scan of the region-of-interest.