Feature Extraction and Classification of Overlapping Cervical Cancer Cells using Convolutional Neural Networks

Conference: McGraw-Hill International Conference on Signal, Image Processing Communication and Automation
Author(s): Seema Singh, Aaron Joseph, Elsa Mathew, A Veena, Deepak Yadav Year: 2017
Grenze ID: 02.MH-ICSIPCA.2017.1.28 Page: 187-190

Abstract

Cervical Cancer is the fourth most common cause of cancer and death in women [1]. India has one of the largest\nnumbers of cases of cervical cancer and efforts are made to reduce the number of fatalities by improving the techniques of\ndetection. The steps that are followed includes obtaining the microscopic images by Liquid Based Cytology(LBC) method,\nfollowed by the classification of the visual input. For this method of classification convolutional neural network is used\nwhich is a very powerful technology for the classification of the visual input arising from the detection process. The\nconvolutional neural network used here is the Inception v3. The end-result is the accurate classification of overlapping as\nwell as non-overlapping cervical cancer cells.

<< BACK

MH-ICSIPCA - 2017