STRAMPN-Histopathological Images for Ovarian Cancer Prediction

Citation Author(s):
National Institute of Technology Hamirpur
Tej Pratap
National Institute of Technology Hamirpur
National Institute of Technology Hamirpur
National Institute of Technology Hamirpur
Kumari Maurya
King George's Medical University
Kumari Verma
Dr B R Ambedkar National Institute of Technology Jalandhar
Nagendra Pratap
Dr B R Ambedkar National Institute of Technology Jalandhar
Submitted by:
Samridhi Singh
Last updated:
Fri, 03/17/2023 - 09:22
Data Format:
0 ratings - Please login to submit your rating.


Ovarian cancer is among the top health issues faced by women everywhere in the world . Ovarian tumours have a wide range of possible causes. Detecting and tracking down these cancers in their early stages is difficult which adds to the difficulty of treatment. In most cases, a woman finds out she has ovarian cancer after it has already spread. In addition, as technology in the field of artificial intelligence advances, detection can be done at an earlier level. Having this data will assist the gynaecologist in treating these tumours as soon as possible. However, a major obstacle to detection is the lack of useful datasets freely available online. For ovarian cancer to be detected, there needs to be a reliable dataset right away. This dataset was made from a collection of histopathological whole slide images found at the source [1]. The whole slide images from the online dataset [1] made it difficult to distinguish between tumours and non-tumorous images as the regions of both were mixed into one slide image. So, our proposed dataset, STRAMPN, is made up of images that were cut from that large slide image and put into two folders called Ovarian_Cancer and Ovarian_Non_Cancer. This was done under the supervision of experienced pathologist and gynaecologist. These images will be easier to process and inspect for research and development purposes in the domain of ovarian cancer detection. The name STRAMPN has derived from the initials of all the authors’ names (Samridhi Singh, Tej Pratap Yadav, Rishabh Dhenkawat, Aryan Verma, Malti Kumari Maurya, Prem Kumari, and Nagendra Pratap Singh respectively) that contributed in creating the dataset. The dataset is constructed for binary classification and analysis. There are 481 images belonging to cancerous tissues in the Ovarian_Cancer category folder and 506 images belonging to the healthy tissues in the Ovarian_Non_Cancer folder.

1. Wang, C.-W., Chang, C.-C., Lo, S.-C., Lin, Y.-J., Liou, Y.-A., Hsu, P.-C., Lee, Y.-C., & Chao, T.-K. (2021). A dataset of histopathological whole slide images for classification of Treatment effectiveness to ovarian cancer [Data set]. The Cancer Imaging Archive.


The dataset contains histopathological images cropped from a dataset of histopathological whole slide images for classification of treatment effectiveness to ovarian cancer. The STRAMPN is the primary file for this STRAMPN dataset which when unzipped, reveals Ovarian_Cancer and Ovarian_Non_Cancer sub directories. The Ovarian_Cancer folder has 481 cancerous tissues images, whereas the Ovarian_Non_Cancer folder contains 506 normal tissue images. There are a total of 987 pictures, and they're all in .jpg format.


Ovarian Cancer IEEE Data Porta 04032024

Submitted by Shahid Khan on Wed, 04/03/2024 - 19:43