Cervical Cell Image Database

This repository contains cervical cell (CC) microscopy images, nucleus segmentation images, and classification data developed for deep learning–based cervical cell classification research. Accurate classification of cervical cells is an important task in computer-aided diagnosis systems for cervical cancer screening. Automated classification methodologies may assist cytopathologists by improving diagnostic efficiency and reducing subjectivity during microscopic evaluation. Repository Contents Nucleus Segmentation Images Contains nucleus segmentation images employed for CNN-based cervical cell classification into normal and abnormal categories. The database includes segmented images derived from the Original Herlev Dataset and contains the following subclasses: Cervical Cell Category Number of Images Carcinoma in situ 150 Light dysplastic 182 Moderate dysplastic 146 Severe dysplastic 197 Normal columnar 98 Normal intermediate 70 Normal superficial 74 Total images: Normal cells: 242 Abnormal cells: 675 Original Dataset Source This work is based on the Original Herlev Dataset described in: Martin, E. Pap-smear Classification, Master’s Thesis, Technical University of Denmark, 2003. Marinakis, Y.; Marinaki, M.; Dounias, G. Particle swarm optimization for pap-smear diagnosis. Expert Systems with Applications, 2008, 35(4), 1645–1656. Tsakonas, A.; Dounias, G.; Jantzen, J.; Axer, H.; Bjerregaard, B.; von Keyserlingk, D.G. Evolving rule-based systems in two medical domains using genetic programming. Artificial Intelligence in Medicine, 2004, 32(3), 195–216.