CNN Coronary Artery Segmentation via Jerman Filter Preprocessing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current image processing systems are unable to efficiently perform autonomous segmentation of contrast-filled coronary artery vessels in x-ray images, which is crucial for accurate diagnosis and monitoring of coronary artery diseases.

Innovation Solution

A computer-implemented method using a convolutional neural network (CNN) with preprocessing steps like the Jerman filter and data augmentation, trained on x-ray angiography scans with binary masks to extract and segment coronary vessels, improving segmentation accuracy through residual connections and optimized training procedures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional image processing systems are used for coronary artery vessel segmentation, then the system complexity remains low, but the segmentation accuracy and reliability are insufficient

Engineering Contradiction:
Improvesegmentation accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces traditional mechanical image processing methods with a neural network-based system. The neural network automatically learns features and performs segmentation, substituting manual feature extraction and traditional algorithms with an intelligent system that achieves superior accuracy while handling the complexity internally.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent transforms the segmentation problem by changing the approach from fixed thresholding and manual parameter tuning to adaptive parameter learning. The neural network automatically adjusts segmentation parameters based on learned patterns from training data, improving accuracy without requiring explicit parameter specification by users.

Inventive Principle:
Principle #35Parameter changes

2Productivity

If manual segmentation methods are used, then the system remains simple, but the productivity and time efficiency are low

Engineering Contradiction:
Improvesegmentation speedVSAvoidprocessing time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent implements preliminary action through an offline training phase where the neural network learns from annotated training data. This pre-learning process enables the system to perform rapid automated segmentation during actual use, significantly improving productivity while reducing the time required for each segmentation task compared to manual methods.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If traditional segmentation algorithms are applied, then false positives occur frequently, but using complex neural networks may increase computational resource consumption

Engineering Contradiction:
Improvereduction of false positivesVSAvoidcomputational resource consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent applies segmentation by dividing the coronary artery segmentation task into distinct neural network components: a U-Net architecture for vessel segmentation and a separate refinement network for boundary optimization. This modular approach improves reliability by addressing different aspects of segmentation separately while managing computational resources efficiently through specialized processing for each component.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3719744B1Method and system for machine learning based segmentation of contrast filled coronary artery vessels on medical images
Publication Date: 2023.08.23 KARDIOLYTICS INC

AI summary

A computer-implemented method for autonomous segmentation of contrast-filled coronary artery vessels, the method comprising the following steps: receiving (101) a x-ray angiography scan representing a maximum intensity projection of a region of anatomy that includes the coronary vessels on the imaging plane; preprocessing (102) the scan to output a preprocessed scan; and performing autonomous coronary vessel segmentation (103) by means of a trained convolutional neural network (CNN) that is trained to process the preprocessed scan data to output a mask denoting the coronary vessels.