Arterial Image Region Classification with Polar-to-Cartesian ML
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OCT imaging systems face challenges in accurately classifying and segmenting arterial tissue and features due to variations in tissue types, leading to errors in visual analysis, particularly in identifying vulnerable plaques and other regions of interest in coronary arteries.
Innovation Solution
A machine learning system utilizing a neural network is trained with annotated arterial image data to classify and segment regions of interest, such as intima, media, and adventitia, using ground truth masks, and can convert polar image data to Cartesian form for accurate feature detection and display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual analysis by cardiologist is used to assess pathological features, then diagnostic accuracy can be maintained, but processing time is excessive and errors occur due to difficulty in discerning certain tissue information
Solution Approach 1:
The patent replaces the manual visual analysis mechanism (cardiologist examination) with an automated machine learning classification system. The ML model processes OCT image data to identify and classify arterial tissue types, plaque characteristics, and vulnerable features automatically, eliminating the time-consuming manual review while maintaining or improving diagnostic accuracy through consistent application of classification criteria.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between the OCT imaging system and the cardiologist. This intermediary automatically performs preliminary classification of tissue regions and highlights suspicious areas, reducing the cognitive load on cardiologists and enabling them to focus on complex decision-making while the routine classification work is handled by the automated system.
2Loss of information
If manual measurement and imaging technologies are used to identify tissue types, then diagnostic information can be obtained, but errors occur because certain information cannot be readily discerned
Solution Approach 1:
The patent transforms the classification approach by changing from manual visual parameter assessment to automated multi-parameter analysis. The machine learning model simultaneously evaluates multiple image features (intensity, texture, shape, spatial relationships) that are difficult for humans to assess manually, extracting comprehensive tissue characteristics and classification results with higher precision and completeness.
3Speed
If OCT imaging is used to image coronary arteries in the beating heart, then real-time imaging capability is achieved, but accurate classification and segmentation of arterial tissue remains challenging due to tissue variations
Solution Approach 1:
The patent employs a dynamic machine learning classification system that adapts to the varying tissue characteristics encountered during rapid OCT imaging of the beating heart. The model processes each frame or small sequence of frames in real-time, adjusting classification parameters based on local tissue variations while maintaining consistent performance across different cardiac phases and anatomical regions.
Data Source
AI summary
In part, the disclosure relates to methods, and systems suitable for evaluating image data from a patient on a real time or substantially real time basis using machine learning (ML) methods and systems. Systems and methods for improving diagnostic tools for end users such as cardiologists and imaging specialists using machine learning techniques applied to specific problems associated with intravascular images that have polar representations. Further, given the use of rotating probes to obtain image data for OCT, IVUS, and other imaging data, dealing with the two coordinate systems associated therewith creates challenges. The present disclosure addresses these and numerous other challenges relating to solving the problem of quickly imaging and diagnosis a patient such that stenting and other procedures may be applied during a single session in the cath lab.