Embedded Files
We developed a guided, semi-automated MATLAB package to process CT images using MATLAB and ITK-SNAP. The designed tool provides means for specialized medical image processing of CT-Scans, laying the foundations for adventitial tissue analysis in coronary atherosclerosis.
We developed a guided, semi-automated MATLAB package to process CT images using MATLAB and ITK-SNAP. The designed tool provides means for specialized medical image processing of CT-Scans, laying the foundations for adventitial tissue analysis in coronary atherosclerosis.
Features of our Design
Features of our Design
Easy and efficient adoption for researchers and clinicians with basic programming knowledge
Normalization of blood vessels through linearization to allow comparisons between different blood vessels, regardless of curvature
Intuitive visualization of different types of tissues through color-coding
Quantitative analysis including the spatial distribution of tissues surrounding atherosclerotic lesions
Extraction of important physiological information useful for atherosclerosis research
3D Vessel Linearization. The left image shows the vessel before we apply our linearization algorithm, and the right image displays the vessel after linearization.
2D Cross-Sectional Visualization. We can move through the entire CT image cross-section by cross-section to view the tissue distribution of the vessel on an extremely localized level. In the images above, red indicates blood, blue indicates the arterial wall, yellow indicates myocardium, purple indicates epicardial fat, and green indicates calcium.
Applications of our Tool
Applications of our Tool
Research: Study the pathophysiology of the disease.
Clinical Trials: Provide information about the efficacy of atherosclerotic therapies.
Physiological Modeling: Create physiological models that resemble the morphology of real coronary vessels and atherosclerotic vessels.
Radiology: Facilitate the identification of atherosclerosis, especially in its earliest stages.
Future directions
Future directions
Improve linearization accuracy
Further verify tissue identification results
Improve user interface and throughput
Apply the pipeline to physiologically relevant studies
End Goal
End Goal
Create a fully automated image processing tool
Utilize our tool in physiologically relevant studies
Generate new insights into the pathology of atherosclerosis
Acknowledgements:
Acknowledgements:
Dr. Geert Schmid-Schoenbein, Dr. Elliot McVeigh, Zhennong Chen, Dr. Bruce Wheeler, and Neha Chhugani
University of California, San Diego
2020
Page Leader: Salil Patel
Page updated
Report abuse