CRIC Searchable Image Database is a platform that enables knowledge discovery from cells, manual labeling, and quantification of cell structures. It allows the creation of collections of specific cell types.
The first collection, named CRIC Cervix, includes 400 real images of Pap smear, 11,384 cells classified in six (6) lesion types.
The CRIC Cervix Classification collection is similar to those obtained during a routine examination, with many cells per image. Cytopathologists classify the cells following the protocol defined by the Bethesda system. In addition to the cells being individually identified and classified manually by different specialists, the whole image also has a diagnosis.
There is a playground platform, an interactive environment where users may add images and use classification and segmentation tools in their own collection. This publicly available web tool provides information about cervical cells, enabling the community to explore cell morphology and variability while facilitating discoveries from the image collections.
The download of CRIC Cervix classification collection is available at CRIC Searchable Image Database. All images are also available as a Figshare collection as CRIC Cervix Cell Classification [1].
Team: Raniere, Caio, Marcus, Mariana, Daniela, Fátima, Claudia and Andrea.
[1] Mariana T. Rezende, Raniere Silva, Fagner de O. Bernardo, Alessandra H. G. Tobias, Paulo H. C. Oliveira, Tales M. Machado, Caio S. Costa, Fatima N. S. Medeiros, Daniela M. Ushizima, Claudia M. Carneiro & Andrea G. C. Bianchi, Cric searchable image database as a public platform for conventional pap smear cytology data. Scientific Data – Nature 8, 151 (2021).
[2] Rezende, Mariana Trevisan; Tobias, Alessandra Hermógenes Gomes; Silva, Raniere; Oliveira, Paulo; Sombra de Medeiros, Fatima; Ushizima, Daniela; et al. (2020): CRIC Cervix Cell Classification. figshare. Collection.
