Identification of coronary artery stenosis based on hybrid segmentation and feature fusion

Title

Identification of coronary artery stenosis based on hybrid segmentation and feature fusion

Creator

Kavipriya K.; Hiremath M.

Description

Coronary artery disease has been the utmost mutual heart disease in the past decades. Various research is going on to prevent this disease. Obstructive CAD occurs when one or more of the coronary arteries which supply blood to myocardium are narrowed owing to plaque build-up on the arteries inner walls, causing stenosis. The fundamental task required for the interpretation of coronary angiography is identification and quantification of severity of stenosis within the coronary circulation. Medical experts use X-ray coronary angiography to identify blood vessel/artery stenosis. Due to the artefact, the image has less clarity and it will be challenging for the medical expert to find the stenosis in the coronary artery. The solution to the problem a computational framework is proposed to segment the artery and spot the location of stenosis in the artery. Here the author presented an automatic method to detect stenosis from the X-ray angiogram image. A unified Computational method of Jerman, Level-set, fine-tuning the artery structure, is developed to extract the segmented artery features and detect the arterys stenosis. The current experimental outcomes illustrate that this computational method achieves average specificity, sensitivity, Accuracy, precision and F-scores of 95%, 97.5%, 98%, 97.5% and 97.5%, respectively. 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.

Source

Automatika, Vol-64, No. 3, pp. 622-633.

Date

2023-01-01

Publisher

Taylor and Francis Ltd.

Subject

Coronary artery disease; Keyframe extraction; stenosis; stenosis detection; X-ray angiography

Coverage

Kavipriya K., Department of Computer Science, Christ Deemed to be University, Bangalore, India; Hiremath M., Department of Computer Science, Christ Deemed to be University, Bangalore, India

Rights

All Open Access; Gold Open Access; Green Open Access

Relation

ISSN: 51144

Format

Online

Language

English

Type

Article

Identifier

https://doi.org/10.1080/00051144.2023.2205727

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85158025936&doi=10.1080%2f00051144.2023.2205727&partnerID=40&md5=c11d2756709585e2124c2f66717c98fb