Anatomical Region Prefetching via Atlas Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inconsistent DICOM tagging across medical imaging systems leads to complex and inefficient exam type identification, delaying deployment of enterprise imaging solutions and causing clinical inefficiencies, especially in large healthcare institutions with numerous exam types.
Innovation Solution
The method involves determining anatomical regions of medical imaging studies through atlas matching and keyword extraction, allowing for automatic categorization independent of DICOM headers, using image analytics and deep learning to reduce manual linking and enhance accuracy, thereby simplifying prefetching processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If inconsistent DICOM tagging is used across medical imaging systems, then system compatibility and ease of integration are improved, but exam type identification accuracy and deployment efficiency deteriorate
Solution Approach 1:
The patent introduces an intermediary anatomical region determination step that mediates between inconsistent DICOM tags and the required exam type identification. By using atlas matching and keyword extraction as intermediate processes, the system translates varied tagging conventions into a unified anatomical region classification, thereby resolving the contradiction between system compatibility and identification accuracy
Solution Approach 2:
The patent replaces the mechanical/manual process of linking exam types with DICOM tags through image analytics and deep learning systems. This substitution enables automatic, accurate determination of anatomical regions and exam types without relying on consistent manual tagging, thus maintaining system compatibility while improving identification accuracy
2Measurement precision
If manual linking of exam types is performed to ensure accuracy, then exam type identification precision is improved, but deployment time and system complexity increase
Solution Approach 1:
The patent implements preliminary action by pre-processing medical imaging studies to extract anatomical region information and generate keyword profiles before the actual exam type identification is needed. This preliminary categorization using atlas matching and deep learning enables rapid, accurate exam type determination during deployment without time-consuming manual linking
Solution Approach 2:
The system performs self-service through automated image analytics and deep learning models that independently determine anatomical regions and classify exam types without human intervention. This self-service capability maintains high identification precision while eliminating the time loss associated with manual processes
3Measurement precision
If comprehensive manual linking of numerous exam types is implemented, then identification accuracy is improved, but device complexity and ease of operation worsen
Solution Approach 1:
The patent extracts the essential anatomical region information from complex DICOM headers and exam descriptions, separating this critical identification data from the overwhelming complexity of comprehensive manual linking. By focusing only on extracted anatomical keywords and atlas-matched regions, the system maintains high identification accuracy while dramatically reducing system configuration complexity
Solution Approach 2:
The patent implements a universal anatomical region classification system that serves multiple exam types and imaging modalities through a single standardized framework. This multi-functional approach replaces the need for separate manual linking configurations for each exam type, thereby maintaining identification accuracy while reducing overall system complexity
4Speed
If traditional prefetching methods are used without anatomical analysis, then processing speed is improved, but prefetching accuracy and clinical relevance deteriorate
Solution Approach 1:
The patent applies preliminary action by pre-determining anatomical regions and generating keyword profiles for medical imaging studies before prefetching decisions are made. This preliminary anatomical analysis enables the system to quickly and accurately identify relevant prior exams during prefetching operations, achieving both high speed and high accuracy without the need for complex real-time analysis
Data Source
AI summary
Inference of appropriate anatomical region from inconsistent descriptions in order to provide fast and accurate prefetching is provided. In various embodiments, an anatomical region of a first medical imaging study is determined. A first plurality of keywords is determined corresponding to the anatomical region of the first medical imaging study. A plurality of studies is accessed having a patient in common with the first medical imaging study. A second plurality of keywords is extracted from the plurality of studies. Those of the plurality of studies having extracted keywords in common with the first plurality of keywords are selected. The selected studies are pre-fetched for display to a user.