DETECTION OF STENOSIS AND OCCLUSION OF CORONARY ARTERIES

Authors

  • Сара Миљуш Autor

DOI:

https://doi.org/10.24867/28RB04Miljus

Keywords:

stenosis and occlusion of coronary arteries, coronary angiography, neural networks, deep learning, YOLOv4 and YOLOv8 algorithms

Abstract

The paper is based on the detection of stenoses and occlusions of the coronary arteries of the heart, as well as finding their percentage of blockage, using machine learning algorithms, namely YOLOv4, YOLOv8 algorithms, and the Detectron2 library. The dataset used for the project consists of 3.850 images of coronary arteries where stenoses and/or occlusions are clearly marked.

References

1. https://epoteka.rs/blog/aortna-stenoza/
2. https://www.ncbi.nlm.nih.gov/books/NBK560507/
3. https://www.stetoskop.info/bolesti-srca-i-krvnih-sudova-kardiologija/koronarna-angiografija
4. http://solair.eunet.rs/~ilicv/neuro.html
5. file:///C:/Users/milju/Downloads/Milosavljevic_Duboko_ucenje%20.pdf
6. https://blog.roboflow.com/a-thorough-breakdown-of-yolov4/
7. https://blog.roboflow.com/whats-new-in-yolov8/
8. https://medium.com/@hirotoschwert/digging-into-detectron-2-47b2e794fabd
9. https://www.v7labs.com/blog/mean-average-precision
10. https://nardus.mpn.gov.rs/bitstream/handle/123456789/21403/Disertacija_13527.pdf?sequence=1

Published

2024-10-09