Localização de Núcleos Celulares em Citologia Oral Usando Métodos de Deep Learning

Access the paper here.

Abstract

Objective: Verify the best approach for the early detection of cancer in the analysis of images from oral cavity slides stained with Papanicolaou. Method: Deep Learning approaches of segmentation and object detection are compared to detect nuclei in images from three different patients’ slides. Result: The best result indicates an IoU of 0.59 for segmentation and 0.81 for object detection. Conclusion: The results show that object detection using the Faster R-CNN model has the potential to be used in conjunction with an image classification model to aid in early cancer detection. The image base and codes used have been made publicly available.

Bibtex

@article{MatiasDetectcaoPAPCBIS2020,
    author={Andr{\'{e}} Vict{\'{o}}ria Matias and Allan Cerentini and Luiz Antonio Buschetto Macarini and Jo{\~{a}}o Gustavo Atkinson Amorim and Felipe Perozzo Dalto{\'{e}} and Aldo von Wangenheim},
    title={Localiza{\c{c}}{\~{a}}o de N{\'{u}}cleos Celulares em Citologia Oral Usando M{\'{e}}todos de Deep Learning},
    year={2020},
    issn={2175-4411},
    publisher = {Journal of Health Informatics},
    url={http://www.jhi-sbis.saude.ws/ojs-jhi/index.php/jhi-sbis/article/view/810}
}