Automatic Coronary Stenosis Detection in Cardiac CT Using PBT Classifier
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual detection of coronary stenosis in cardiac CT images is tedious and prone to inter-observer variability, and existing automatic detection methods are limited to detecting only calcified plaques, lacking effectiveness for non-calcified plaques.
Innovation Solution
A fully automatic learning-based method using a probabilistic boosting tree (PBT) classifier to detect both calcified and non-calcified plaques by classifying control points along estimated coronary artery centerlines in cardiac CT volumes, based on local features and trained with contrast-enhanced data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual detection and segmentation of stenosis in cardiac CT images is performed, then detection capability is achieved, but the process is tedious and subject to inter-observer variability
Solution Approach 1:
The system performs automatic stenosis detection without requiring manual intervention. The computer automatically segments coronary arteries, detects plaques, and classifies stenosis regions using trained classifiers, eliminating the need for manual detection while improving reliability and reducing time loss
Solution Approach 2:
The patent replaces manual mechanical detection processes with automated computational methods. Machine learning classifiers and image processing algorithms substitute for human observers, eliminating inter-observer variability and significantly reducing detection time while maintaining or improving accuracy
2Extent of automation
If existing automatic detection methods are used, then detection automation is achieved, but only calcified plaques can be detected, not non-calcified plaques
Solution Approach 1:
The system is designed to detect multiple types of plaques (both calcified and non-calcified) using a unified automated framework. The classifier is trained to recognize different plaque characteristics, making the system versatile rather than limited to a single plaque type
Solution Approach 2:
The patent changes the detection parameters and features used in the classifier to accommodate different plaque types. By adjusting the training data and feature extraction methods, the system adapts to detect both calcified (bright regions) and non-calcified (darker regions) plaques within the same automated framework
3Measurement precision
If a trained classifier is used to detect stenosis regions, then detection accuracy is improved, but the system complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-training classifiers on labeled data before actual detection. Feature extraction templates and classification models are prepared in advance, allowing the system to achieve high detection accuracy during operation without complex real-time processing
Solution Approach 2:
The patent segments the detection process into distinct stages: coronary artery segmentation, control point identification, feature extraction, and classification. This modular segmentation manages system complexity by breaking down the complex detection task into manageable, independent modules that can be processed sequentially
Data Source
AI summary
A method and system for automatic coronary stenosis detection in computed tomography (CT) data is disclosed. Coronary artery centerlines are obtained in an input cardiac CT volume. A trained classifier, such as a probabilistic boosting tree (PBT) classifier, is used to detect stenosis regions along the centerlines in the input cardiac CT volume. The classifier classifies each of the control points that define the coronary artery centerlines as a stenosis point or a non-stenosis point.