SEGMENTAÇÃO DE ARTÉRIAS CORONÁRIAS EM IMAGENS DE ANGIOGRAFIA POR TOMOGRAFIA COMPUTADORIZADA UTILIZANDO REDES NEURAIS CONVOLUCIONAIS 3D

Esbel Tomas Valero Orellana; Marcelo Ossamu Honda; Paulo Eduardo Ambrósio; Cleviton Borges de Jesus; Allan de Medeiros Martins; Dany Sanchez Dominguez

doi:10.14295/vetor.v34i2.18531

Authors

Esbel Tomas Valero Orellana UNIVERSIDADE ESTADUAL DE SANTA CRUZ https://orcid.org/0000-0002-4087-2722
Marcelo Ossamu Honda Universidade Estadual de Santa Cruz https://orcid.org/0000-0001-9008-2211
Paulo Eduardo Ambrósio Universidade Estadual de Santa Cruz https://orcid.org/0000-0002-6043-3591
Cleviton Borges de Jesus Centro de Pesquisa e Desenvolvimento Tecnológico em Informática e Eletroeletrônica https://orcid.org/0009-0007-4009-3968
Allan de Medeiros Martins Centro de Pesquisa e Desenvolvimento Tecnológico em Informática e Eletroeletrônica de Ilhéus
Dany Sanchez Dominguez Universidade Estadual de Santa Cruz https://orcid.org/0000-0002-0640-2970

DOI:

https://doi.org/10.14295/vetor.v34i2.18531

Keywords:

Coronary artery, Segmentation, Deep Learning, Computer-aided diagnostic, Medical images

Abstract

Coronary Artery Disease (CAD) is the leading cause of death from cardiovascular diseases worldwide. Accurate risk assessment of CAD is crucial for prevention. Computed Tomography Angiography (CTA) is a widely used non-invasive method for diagnosing CAD. Accurate segmentation of coronary arteries in CTA images is essential for quantifying the disease and aiding in diagnosis. In this study, we evaluated a dataset of CTA images provided by the ImageCAS project and compared different segmentation algorithms, including a proposed direct segmentation method. We evaluated the performance of the algorithms using the Dice similarity coefficient, comparing the results to a ground truth. We experimented with different image resolutions to analyze the impact on performance and computational resource consumption. Additionally, we propose an ensemble method to combine the results of different algorithms, aiming to improve segmentation accuracy. The obtained results demonstrate that the proposed ensemble method achieves superior performance compared to individual algorithms.

Downloads

Download data is not yet available.

References

World Heart Federation, “World heart report 2023: Confronting the world’s number one killer,” World Heart Federation, Geneva, Switzerland, Relatório técnico, 2023. Disponível em: https://www.world-heart-federation.org/world-heart-report/

W. Huang, L. Huang, Z. Lin, S. Huang, Y. Chi, J. Zhou, J.-M. Zhang, ru san tan, e L. Zhong, “Coronary artery segmentation by deep learning neural networks on computed tomographic coronary angiographic images,” em Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2018, pp. 608–611. Disponível em: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8512328

A. Colombo e F. Giannini, “Is it time to replace conventional angiography with coronary computed tomography?” European Heart Journal, vol. 39, no. 41, pp. 3699–3700, 2018. Disponível em: https://doi.org/10.1093/eurheartj/ehy578

Q. Xie, L. Zhou, Y. Li, R. Zhang, H. Wei, G. Ma, Y. Tang, e P. Xiao, “Comparison of prognosis between coronary computed tomography angiography versus invasive coronary angiography for stable coronary artery disease: a systematic review and meta-analysis,” Frontiers in Cardiovascular Medicine, vol. 10, p. 1010536, 2023. Disponível em: https://doi.org/10.3389/fcvm.2023.1010536

A. Zeng, C. Wu, G. Lin, W. Xie, J. Hong, M. Huang, J. Zhuang, S. Bi, D. Pan, N. Ullah, K. N. Khan, T. Wang, Y. Shi, X. Li, e X. Xu, “ImageCAS: A large-scale dataset and benchmark for coronary artery segmentation based on computed tomography angiography images,” Computerized Medical Imaging and Graphics, vol. 109, p. 102287, 2023. Disponível em: https://doi.org/10.1016/j.compmedimag.2023.102287

A. Rouhollahi, J. N. Willi, S. Haltmeier, A. Mehrtash, R. Straughan, H. Javadikasgari, J. Brown, A. Itoh, K. I. de la Cruz, E. Aikawa, E. R. Edelman, e F. R. Nezami, “CardioVision: A fully automated deep learning package for medical image segmentation and reconstruction generating digital twins for patients with aortic stenosis,” Computerized Medical Imaging and Graphics, vol. 109, p. 102289, 2023. Disponível em: https://doi.org/10.1016/j.compmedimag.2023.102289

Y. Shen, Z. Fang, Y. Gao, N. Xiong, C. Zhong, e X. Tang, “Coronary Arteries Segmentation Based on 3D FCN with Attention Gate and Level Set Function,” IEEE Access, vol. 7, pp. 42 826–42 835, 2019. Disponível em: https://doi.org/10.1109/ACCESS.2019.2908039

O. Ronneberger, P. Fischer, e T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” em Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, 2015, pp. 234–241, arXiv: 1505.04597. Disponível em: http://arxiv.org/abs/1505.04597

C. Huang, H. Han, Q. Yao, S. Zhu, e S. K. Zhou, “3D U-Net: A 3D Universal U-Net for Multi-Domain Medical Image Segmentation,” em Medical Image Computing and Computer Assisted Intervention – MICCAI 2019, 2019, pp. 291–299, arXiv: 1909.06012. Disponível em: http://arxiv.org/abs/1909.06012