Predição de lesões celulares em imagens de citologia convencional usando redes neurais convolucionais

• Lucas A. Freitas UFOP
• Débora N. Diniz UFOP
• Marcone J. F. Souza UFOP
• Cláudia M. Carneiro UFOP
• Daniela M. Ushizima Lawrence Berkeley National Laboratory
• Fátima N. S. de Medeiros UFC
• Andrea G. C. Bianchi UFOP

DOI: https://doi.org/10.5753/sbcas.2023.229938

Abstract

This article presents a novel methodology based on deep learning for detecting cervical lesions in Pap smear samples. The proposed model uses nucleus location information and performs cropping around it using different dimensions, without the need for image segmentation. Several CNN models were developed and trained using real cervical cell images. The results showed that the model achieved a satisfactory accuracy of 0.94 using a box size of 70x70 without needing image segmentation. The proposed methodology can assist cytopathologists in improving diagnosis and laboratory results quality, contributing to cervical cancer prevention.

References

Diniz, D., Keller, B., Rezende, M., Bianchi, A., Carneiro, C., Oliveira, R., Luz, E., Ushizima, D., Medeiros, F., and Souza, M. (2022). A cytopathologist eye assistant for cell screening. AppliedMath, 2:659–674.

Diniz, D. N., Rezende, M. T., Bianchi, A. G. C., Carneiro, C. M., Luz, E. J. S., Moreira, G. J. P., Ushizima, D. M., de Medeiros, F. N. S., and Souza, M. J. F. (2021a). A deep learning ensemble method to assist cytopathologists in pap test image classification. Journal of Imaging, 7(7):111.

Diniz, D. N., Rezende, M. T., Bianchi, A. G. C., Carneiro, C. M., Ushizima, D. M., de Medeiros, F. N. S., and Souza, M. J. F. (2021b). A hierarchical feature-based methodology to perform cervical cancer classification. Applied Sciences, 11(9).

Gómez, O. H., Sánchez-DelaCruz, E., and de la Mata, A. P. (2017). Classification of cervical cancer using assembled algorithms in microscopic images of papanicolaou. Res. Comput. Sci., 139:125–134.

Isidoro, D. W. A. (2021). Classificação automática de recortes de células cervicais baseada em características não geométricas. Trabalho de Conclusão de Curso (Ciência da Computação), Departamento de Computação, Universidade Federal de Ouro Preto, Ouro Preto, MG. Disponível em: [link].

Mariarputham, E. J. and Stephen, A. (2015). Nominated texture based cervical cancer classification. Comput Math Methods Med, 2015:586928.

Pangarkar, M. A. (2022). The bethesda system for reporting cervical cytology. Cytojournal, 19:28. PMCID: PMC9168399.

Rezende, M. T., Bianchi, A. G., and Carneiro, C. M. (2021). Cervical cancer: Automation of pap test screening. Diagnostic Cytopathology, 49(4):559–574.

Sung, H., Ferlay, J., Siegel, R. L., Laversanne, M., Soerjomataram, I., Jemal, A., and Bray, F. (2021). Global cancer statistics 2020: Globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: a cancer journal for clinicians, 71(3):209–249.

Who (2022). World Health Organization Cervical Cancer. Disponível em: [link]. Acessado em: 20-03-2023.