Liver MRI Hepatobiliary Phase Prediction for Shorter Exams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MRI examinations of the liver using hepatobiliary contrast agents are lengthy due to the need for multiple scans across different phases, causing patient discomfort from prolonged immobilization to avoid motion artifacts.
Innovation Solution
A machine learning model is trained to predict an MRI image of the liver in the hepatobiliary phase based on a single initial MRI scan in the transition phase, optionally combined with a native scan without contrast, eliminating the need for additional scans.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional MRI examination protocols are used for liver imaging, then comprehensive diagnostic information can be obtained, but the examination time becomes excessively long (15-30 minutes or more)
Solution Approach 1:
The examination protocol is segmented into multiple groups where each group targets specific liver zones (e.g., left lobe, right lobe, caudate lobe) with dedicated imaging sequences. This allows the system to acquire comprehensive diagnostic information across all liver regions while significantly reducing total examination time compared to conventional full-liver imaging protocols.
2Measurement precision
If multiple imaging sequences are acquired for different liver zones, then diagnostic accuracy is improved, but the complexity of the examination protocol increases
Solution Approach 1:
The examination protocol dynamically adapts to patient-specific anatomy and clinical indications. The system automatically adjusts which liver zones require imaging and selects appropriate sequences based on real-time assessment, transforming a static complex protocol into a dynamic, streamlined process that maintains diagnostic accuracy while reducing unnecessary imaging steps.
Solution Approach 2:
Different imaging sequences and parameters are applied to different liver zones based on their specific diagnostic requirements. For example, certain zones may require contrast-enhanced sequences while others use non-contrast imaging, allowing optimized diagnostic accuracy for each region without uniformly applying complex protocols throughout the entire liver.
3Measurement precision
If the patient remains stationary during imaging, then image quality is maintained, but patient comfort deteriorates and motion artifacts increase
Solution Approach 1:
The imaging protocol employs periodic acquisition patterns with brief pauses between sequences, allowing patients to adjust their position slightly without compromising overall image quality. This periodic rhythm maintains patient comfort while capturing diagnostic information at critical phases of the imaging cycle.
Data Source
Figure 1(a)~1(f)
Figure 2
Figure 3
AI summary
The present invention is concerned with the acceleration of an MRI examination of the liver by means of a machine learning model. The machine learning model is configured and trained to predict an MRI image of a liver in the hepatobiliary phase after the application of a hepatobiliary contrast agent on the basis of one or more MRI images which were produced at a time in an earlier phase. The present invention relates to a method for training the machine learning model, to a computer-implemented method for predicting the MRI image in the hepatobiliary phase by means of the trained model, and to a computer system and a computer program product for carrying out the prediction method.