Automatic detection and visualization system for coronary artery calcification using optical frequency domain imaging

Percutaneous coronary intervention (PCI) is mainly used in the treatment of stenosis of the coronary arteries of the heart characteristic of coronary artery disease, and it is important that the level of calcification is evaluated in advance of this procedure. A physician typically examines cross-se...

Full description

Saved in:
Bibliographic Details
Published in:Artificial life and robotics 2023-05, Vol.28 (2), p.460-470
Main Authors: Oikawa, Ryo, Doi, Akio, Ishida, Masaru, Chakraborty, Basabi
Format: Article
Language:English
Subjects:
Arteries
Artificial Intelligence
Artificial neural networks
Calcification
Cardiovascular disease
Computation by Abstract Devices
Computer Science
Control
Coronary artery disease
Coronary vessels
Frequency domain analysis
Image segmentation
Machine learning
Mechatronics
Optical frequency
Original Article
Physicians
Robotics
Semantic segmentation
Visualization
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Percutaneous coronary intervention (PCI) is mainly used in the treatment of stenosis of the coronary arteries of the heart characteristic of coronary artery disease, and it is important that the level of calcification is evaluated in advance of this procedure. A physician typically examines cross-sectional OFDI images of the coronary artery and decides whether PCI is applicable. However, it takes a lot of time to interpret many sliced images. It is difficult to accurately assess the entire calcified area from the individual slices. To solve these problems, we propose an automatic detection and visualization system for coronary artery calcification by using images obtained from optical frequency domain imaging (OFDI). This system assists physicians by automatically detecting and intuitively visualizing calcified areas in a short period of time. The system is built using DeepLabv3+ , a deep-learning network for semantic segmentation. The deep neural network was trained using 2149 coronary OFDI images labeled by physicians.
ISSN:1433-5298
1614-7456
DOI:10.1007/s10015-023-00854-2