Nucleus Detection in Cervical Samples Stained With AgNOR
Published in Computer on the Beach 2022 - COPT XIII, 2022
Abstract
Cervical cancer is a public health problem, where the treatment has a better chance of success if detected early. This paper explores one way of to analyze argyrophilic nucleolus organizer regions (AgNOR) stained slide using deep learning approaches of object detection for detecting the different categories of nucleus. Our results show that a balanced dataset between the explored categories was essential, also that a ResNet-50 as backbone of Fast RCNN shows an AP of 61.8% and 42.5% to detect nucleus and out of focus nucleus.
Bibtex
@inproceedings{AtkinsonDetectionAgNORCOTB2022,
author={Jo{\~{a}}o Gustavo Atkinson Amorim and Vin{\'{i}}us Moreno Sanches and Tainee Bottamedi and Andr{\'{e}} Vict{\'{o}}ria Matias and Marco Cavaco and Alexandre Sherlley Casimiro Onofre and Fabiana Botelho De Miranda Onofre and Aldo von Wangenheim},
title={Nucleus detection in cervical samples stained with AgNOR},
year={2022},
month = {jul},
booktitle = {Anais do {XIII} Computer on the Beach - {COTB}{\textquotesingle}22}},
pages={045-050},
doi={10.14210/cotb.v13.p045-050},
url={http://dx.doi.org/10.14210/cotb.v13.p045-050},
publisher = {Universidade do Vale do Itaja{\'{\i}}}
}