Radiomic Feature’s Selection and Harmonization of COVID-19 in Computed Tomography
DOI:
https://doi.org/10.29384/rbfm.2022.v16.19849001670Keywords:
computed tomography, COVID-19, radiomics, harmonizationAbstract
The advent of radiomics has emerged with opportunities in various diagnostic applications, including chest computed tomography (CT) for diagnosis and staging of COVID-19. This work demonstrates radiomic selection and harmonization strategies in a multicenter study CT of COVID-19. 37 patients with COVID-19 and 36 patients with other pneumonias not associated with COVID-19 were selected from three diagnostic centers. Ninety-four radiomic features were quantified and selected using two methods: 1) supervised, using univariate analysis by the ROC curve, and 2) unsupervised, using intraclass observer agreement (ICC > 0.9) from CT replicates with variations image quality and segmentation. Additionally, population characteristics that could bias feature’s harmonization were analyzed. The unsupervised selection method proved to be effective in the screening of harmonized features, without the need for numerical harmonization. Features selected by the supervised method had larger discrepancy among scanner distributions leading to the need of ComBat harmonization after evaluation of ecological bias sensitivity.
Downloads
References
Brasil. Diretrizes Brasileiras para Tratamento Hospitalar do Paciente com COVID-19. Ministério da Saúde. [S.l.]: [s.n.], 2021, p. 1–53.
Zu, Z. Y.; Jiang, M.; Xu, P. P.; Chen, W.; Ni, Q. Q.; Lu, G. M.; Zhang, L. J. Coronavirus Disease 2019 (COVID-19): A Perspective from China. Radiology, 2020. v. 296, n. 2, p. E15–E25.
Qingxia Wu, Shuo Wang, Liang Li, Qingxia Wu, Wei Qian, Yahua Hu, Li Li, Xuezhi Zhou, He Ma, Hongjun Li, Meiyun Wang, Xiaoming Qiu, Yunfei Zha and Jie Tian. Radiomics Analysis of Computed Tomography helps predict poor prognostic outcome in COVID-19. Theranostics 2020; 10(16): 7231-7244.
Machado, M. A. D.; Pereira Pita, J. L.; Martins Netto, E. PET Radiomic Features Variability: A Phantom Study on the Influence of Reconstruction and Discretization Method. Revista Brasileira de Física Médica, 2021. v. 15, p. 598.
Orlhac, F.; Boughdad, S.; Philippe, C.; Bourdillon, H.; Nioche, C.; Champion, L.; Soussan, M.; Frouin, F.; Frouin, V.; Buvat, I. A postreconstruction harmonization method for multicenter radiomic studies in PET. Journal of Nuclear Medicine, 2018. v. 59, n. 8, p. 1321–1328.
Orlhac, F.; Frouin, F.; Nioche, C.; Ayache, N.; Buvat, I. Validation of a method to compensate multicenter effects affecting CT radiomics. Radiology, 2019. v. 291, n. 1, p. 53–59.
Machado, M. A. D.; Menezes, V. O.; Namías, M.; Vieira, N. S.; Queiroz, C. C.; Matheoud, R.; Alessio, A. M.; Oliveira, M. L. Protocols for harmonized quantification and noise reduction in low-dose oncologic 18F-FDG PET/CT imaging. Journal of Nuclear Medicine Technology, 2019. v. 47, n. 1, p. 47–54.
D. P. Fan et al., "Inf-Net: Automatic COVID-19 Lung Infection Segmentation from CT Images," in IEEE Transactions on Medical Imaging, vol. 39, no. 8, pp. 2626-2637, Aug. 2020, doi: 10.1109/TMI.2020.2996645.
Van Griethuysen, J. J., Fedorov, A., Parmar, C., Hosny, A., Aucoin, N., Narayan, V., Aerts, H. J. (2017). Computational radiomics system to decode the radiographic phenotype. Cancer research, 77(21), e104-e107.
Mukaka, M. M. 81576-194640-1-Pb. Malawi Medical Journal, 2012. v. 24, n. September, p. 69–71. Disponível em: <https://www.ajol.info/index.php/mmj/article/view/81576>.
Johnson, W. E.; Li, Cheng; Rabinovic, A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics, 2007. v. 8, n. 1, p. 118–127.
Fortin, J. P. et al. Harmonization of cortical thickness measurements across scanners and sites. NeuroImage, 2018. v. 167, n. June 2017, p. 104–120. Disponível em: <https://doi.org/10.1016/j.neuroimage.2017.11.024>.
Downloads
Additional Files
Published
How to Cite
Issue
Section
License
Copyright (c) 2022 Danillo Menezes Oliveira, Marcos Antônio Dórea Machado, Mércia Liane de Oliveira, Ronnyldo Rodrigues Eloy da Silva, Mauro Namías, Thamiris Rosado Reina, Paulo César de Oliveira Affonso Júnior, Andreia de Seixas Lessa, Arquimedes Aires Braga de Lira, Krsna Murari de Albuquerque Rodrigues, Thauan Fernandes Moraes, Vinícius de Oliveira Menezes
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
The submission of original articles to the Brazilian Journal of Medical Physics implies the transfer, by the authors, of the rights of print and digital publication. Copyright for published articles remains with the author, with journal rights on first publication. Authors may only use the same results in other publications by clearly indicating this journal as the original publisher. As we are an open access journal, free use of articles in educational, scientific, non-commercial applications is allowed, as long as the source is cited.
The Brazilian Journal of Medical Physics is under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0).
