DICOM Summary Image Generation for CT Anomaly Review
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack efficient methods for detecting pathological anomalies in DICOM images and generating summary images that can be integrated into DICOM files for review in PACS systems, which are essential for accurate diagnosis and analysis in medical imaging.
Innovation Solution
A system utilizing a trained machine learning model to detect pathological anomalies in CT images, generate a summary graphic image with sections indicating anomaly presence, location, and confidence levels, and insert this summary into a DICOM file for review in PACS systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual review of DICOM images is performed by radiologists, then diagnostic accuracy is maintained, but time consumption and workload increase significantly
Solution Approach 1:
The patent introduces an automated anomaly detection system as an intermediary between the DICOM images and the radiologist's review process. The system generates summary images that highlight detected anomalies and provide preliminary analysis, serving as a mediator that prepares and prioritizes cases for radiologist review, thereby reducing their workload and time consumption while maintaining diagnostic accuracy.
Solution Approach 2:
The system performs preliminary analysis of DICOM images by automatically detecting anomalies and generating summary images before radiologist review. This preliminary action identifies and flags potential issues, allowing radiologists to focus their expertise on confirmed or complex cases rather than performing initial screening on all images, thus improving overall diagnostic efficiency.
2Measurement precision
If comprehensive anomaly detection is performed across all DICOM images, then detection accuracy improves, but processing complexity and computational resources increase
Solution Approach 1:
The patent extracts and highlights only the most relevant information from comprehensive DICOM image analysis by generating summary images that focus on detected anomalies. Rather than presenting all processed information, the system extracts key findings and presents them in a condensed visual format, maintaining detection accuracy while reducing the complexity of information presentation and interpretation.
3Loss of information
If detailed anomaly information is provided in all sections of the summary image, then diagnostic completeness improves, but information overload and interpretation difficulty increase
Solution Approach 1:
The summary image employs local quality by differentiating the presentation of information across different sections. Detected anomalies are highlighted with visual indicators and detailed information, while normal findings are presented more concisely or with less emphasis. This localized differentiation ensures complete diagnostic information is available while making the most critical information stand out for easier interpretation.
Solution Approach 2:
The summary image is segmented into distinct sections: abnormality presence indication, detected findings, findings location, and undetected findings. This segmentation organizes comprehensive diagnostic information into manageable, purpose-specific areas, allowing radiologists to quickly locate and interpret relevant information without being overwhelmed by a monolithic presentation of all data.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media for pathological anomaly detection in CT scan images and generation of a summary graphic image. The generated summary graphic image includes a abnormality presence indication section, a detected findings section, a findings location section, and an undetected findings section. The generated summary graphic image is inserted into a DICOM file that is viewable by a software system interacting with a picture archiving and communication system (PACS).