Dataset of Cervical Cell Images for the Study of Changes Associated with Malignancy in Conventional Pap Test

This data set was approved by CEP, a human-research ethics committee {Comitê de Ética em Pesquisa de Campinas}, Brazil (CAAE approval number: 71277217.6.0000.5404) Cervical cancer prevention campaign was carried out at the University of Campinas (Brazil), where 71 women were treated. For this study, six women were chosen: non-pre-menopausal, non-pregnant and between the ages of 25 and 40 years. Samples diagnosed with normal squamous lesions, atypia in squamous cells of undetermined significance (AC-US), low-grade squamous intraepithelial lesion LEI (changes associated with HPV infection or light dislocation (NIC 1)), and immature squamous metaplasia, were chosen. All samples were collected before treatment. The patients were anonymized and assigned a unique identifier. The criterion to limit the number of plates used in this study is related to the large amount of data to be processed and the computational cost necessary for this process. Raw data sets are available as CSV files. Each data was numbered to include the identification of the meta-data, along with the tag of the images following a sequence. For example, the data file “ 01-CAP091868-MORF.csv ” has the prefix 01 that refers to the sequential numbering of the files with the data from the digitized slide. The data set contains 102 digitized images of the human papilloma examination of 7 different patients. They were segmented 962