Artificial Intelligence-Assisted Intravascular Ultrasound Examination in Coronary Artery Disease
A systematic review examines the impact of artificial intelligence and deep learning algorithms on intravascular ultrasound analyses in the treatment of coronary artery disease.
A systematic review evaluated the role of artificial intelligence, machine learning, and deep learning algorithms in the automated analysis and interpretation of intravascular ultrasound examinations used in coronary artery disease.
Coronary Artery Disease and Imaging
Coronary artery disease is characterized by the accumulation of atherosclerotic plaques in the arterial walls and ranks among the leading causes of death and morbidity globally.
Although conventional coronary angiography provides a standard guidance technique, it offers limited information regarding plaque burden and vessel wall architecture.
The Role of Intravascular Ultrasound
Intravascular ultrasound largely eliminates these limitations by providing cross-sectional images of the vessel wall.
However, the interpretation of conventional data is operator-dependent and time-consuming.
Artificial Intelligence and Automation Technologies
Recent developments incorporate artificial intelligence algorithms to perform tasks such as lumen and vessel boundary detection, plaque characterization, and tissue classification.
These technologies aim to simplify catheterization laboratory workflows, reduce observer variability, and improve procedure planning.
Systematic Review Method and Scope
This systematic review was conducted in accordance with the PRISMA 2020 statement and registered in the PROSPERO database.
Eligible studies consisted of original peer-reviewed research involving human coronary datasets and evaluating algorithms for image analysis.
Database Searches and Criteria
Searches were conducted in PubMed and Cochrane CENTRAL databases, filtered for English language and human studies.
The PICO framework aimed to evaluate artificial intelligence interventions to measure diagnostic accuracy, segmentation performance, and processing time.