Pap Smear Cervical Cell Nuclei Segmentation Database

This repository contains Pap smear microscopy images, ground truth annotations, mean-shift segmentation results, and nuclei identification outputs developed for cervical cell nuclei segmentation and classification research in images containing multiple overlapping cells.

Repository Contents

The repository contains:

1. Original Pap Smear Images

Original cervical cytology images provided by Plissiti et al.

JPEG image format.

2. Ground Truth Images

Contains:

nuclei localization annotations,

segmentation ground truth,

expert cytologist identification data.

3. Mean-Shift Segmentation Results

Contains:

oversegmentation outputs,

candidate nuclei region generation results.

4. Nuclei Segmentation Results

Contains:

final nuclei segmentation outputs,

nuclei identification results,

classification visualisation outputs.

Dataset Characteristics

Number of images:

252 images

Image format:

JPEG

Image dimensions:

340 × 256 pixels

Original source images:

Seven high-resolution images

Resolution: 2040 × 1536 pixels

Magnification: 10×

Total nuclei analysed:

711 nuclei

Dataset Source

The original high-resolution Pap smear images were kindly provided by:

Plissiti, M.E.; Nikou, C.; Charchanti, A.Automated Detection of Cell Nuclei in Pap Smear Images Using Morphological Reconstruction and Clustering.IEEE Transactions on Information Technology in Biomedicine, 2011, Vol. 15, pp. 233–241.

The images provided in this dataset were generated in:

TELLO-MIJARES, Santiago; BESCOS, Jesus; FLORES, Francisco.Nuclei Segmentation and Identification in Practical Pap-Smear Images with Multiple Overlapping Cells.Journal of Medical Imaging and Health Informatics, 2016, Vol. 6, No. 4, pp. 992–1000.