Anchor-Prior Image Matching for Cross-Modality Study Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current rule-based engines in medical imaging systems are unreliable for identifying related images across different imaging modalities due to variations in DICOM data nomenclature, leading to inefficient and time-consuming manual searches for relevant prior studies.
Innovation Solution
A computer-implemented method using machine learning models to automatically identify optimal anchor-prior image pairs by extracting features from new and historical medical imaging data, enabling accurate matching without extensive user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If rule-based engines are used to identify related images based on DICOM metadata, then the system can operate with simple matching logic, but the reliability of identifying related images across different imaging modalities deteriorates due to variations in DICOM data nomenclature
Solution Approach 1:
The patent replaces the mechanical rule-based matching system with a machine learning-based semantic understanding system. The ML model learns to interpret the meaning behind DICOM metadata and image content, enabling reliable matching across different modalities and nomenclatures without requiring complex manual rule sets for each scenario
Solution Approach 2:
The patent transforms the matching approach from exact parameter matching (rule-based) to semantic parameter understanding (ML-based). Instead of requiring identical DICOM tags and values, the system learns to recognize semantically equivalent representations across different imaging modalities and vendors through trained models
2Adaptability or versatility
If manual searching and viewport manipulation are performed to find relevant prior studies, then the system can handle diverse imaging scenarios, but the time required for the review process increases significantly
Solution Approach 1:
The patent performs preliminary actions by pre-processing and extracting semantic features from images and metadata in advance. The machine learning models are trained beforehand to understand relationships between images, so when a radiologist needs to compare studies, the system has already prepared the semantic representations needed for rapid matching
Solution Approach 2:
The system performs self-service by automatically identifying and retrieving relevant prior studies without requiring manual searching. The machine learning model autonomously understands the radiologist's needs based on the anchor image and automatically configures the viewport layout, eliminating the need for manual manipulation
3Ease of operation
If radiologists manually configure display protocols and manipulate images in viewports, then the display can be customized for specific needs, but the efficiency and speed of the review process deteriorates
Solution Approach 1:
The system provides self-service by automatically configuring display protocols based on the semantic understanding of the anchor image and retrieved prior studies. The machine learning model determines the optimal viewport layout and image placements without requiring radiologist intervention, thus maintaining ease of operation while dramatically improving productivity
Data Source
AI summary
Systems, methods, and apparatuses implementing a display optimization system are provided herein. In some embodiments, an example display optimization system may be configured to perform an optimal anchor-prior matching operation to identify optimal anchor-prior image pairs or series pairs from new medical imaging data (e.g., one or more anchor image series) and historical medical imaging data (e.g., one or more prior image series).