Detecção de núcleos em imagens citológicas de AgNOR utilizando Aprendizado Profundo

Access the paper here.

Abstract

Evaluate the use of object detection models for the detection of nuclei stained with silver from patients with cervical cancers. Method: Using images of a dataset containing slides stained with silver nitrate, a method known as AgNOR. However, this method of cytology has not yet been much explored, mainly using computational methods. Thus, the Faster-RCNN model is compared with different backbones for the detection of nuclei in images stained by the AgNOR method. Result: The work reached the values of 0.66, 0.79, and 0.80 for accuracy, precision, and recall respectively using a Faster-RCNN model. Conclusion: Faster-RCNN can detect individual nuclei stained with the AgNOR method.

Bibtex

@inproceedings{AtkinsonDetectcaoAgNORCBIS2020,
    author={Jo{\~{a}}o Gustavo Atkinson Amorim and Luiz Antonio Buschetto Macarini and Andr{\'{e}} Vict{\'{o}}ria Matias and Allan Cerentini and Fabiana Botelho De Miranda Onofre and Alexandre Sherlley Casimiro Onofre and Aldo von Wangenheim},
    title={Detec{\c{c}}{\~{a}}o de n{\'{u}}cleos em imagens citol{\'{o}}gicas de AgNOR utilizando Aprendizado Profundo},
    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/811}
}