DCAVN: Cervical cancer prediction and classification using deep convolutional and variational autoencoder network
- PDF / 1,566,468 Bytes
- 17 Pages / 439.37 x 666.142 pts Page_size
- 73 Downloads / 172 Views
DCAVN: Cervical cancer prediction and classification using deep convolutional and variational autoencoder network Aditya Khamparia 1 & Deepak Gupta 2 & Joel J. P. C. Rodrigues 3,4 & Victor Hugo C. de Albuquerque 5 Received: 2 March 2020 / Revised: 15 July 2020 / Accepted: 12 August 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Early detection, early diagnosis and classification of the cancer type facilitates faster disease management of patients. Cervical cancer is fourth most pervasive cancer type which affects life of many people worldwide. The intent of this study is to automate cancer diagnosis and classification through deep learning techniques to ensure patients health condition progress timely. For this research, Herlev dataset was utilized which contains 917 benchmarked pap smear cells of cervical with 26 attributes and two target variables for training and testing phase. We have adopted combination of convolutional network with variational autoencoder for data classification. The usage of variational autoencoder reduces the dimensionality of data for further processing with involvement of softmax layer for training. The results have been obtained over 917 cancerous image type pap smear cells, where 70% (642) allocated for training and remaining 30% (275) considered for test data set. The proposed architecture achieved variational accuracy of 99.2% with 2*2 filter size and 99.4% with 3*3 filter size using different epochs. The proposed hybrid variational convolutional autoencoder approach applied first time for cervical cancer diagnosis and performed better than traditional machine learning methods. Keywords Variational . Convolution . Cervical . Deep learning . Autoencoder
* Aditya Khamparia [email protected]
1
School of Computer Science and Engineering, Lovely Professional University, Punjab, India
2
Maharaja Agrasen Institute of Technology, Rohini, New Delhi, India
3
Federal University of Piaui, Teresina, PI, Brazil
4
Instituto de Telecommunicações, Lisbon 1049-001, Portugal
5
Graduate Program in Applied Informatics, University of Fortaleza, Fortaleza, Brazil
Multimedia Tools and Applications
1 Introduction As per recent reports generated by International Cancer agency, every year around 14 million cases of cancer get detected and 8 million people die due to cancer [5]. In contrast to developed countries, developing country like India alone covers one quarter of the burden of cervical cancer worldwide. According to reports of World Health Organization (WHO), cervical cancer growing rapidly among Indian Women which occurs approximately 1 in 53 women as compared to 1 in 100 women suffering from such ailment worldwide [9]. Further data reported by WHO investigated that usually 67,477 women die from that disease and 122,844 women are diagnosed in different medical care centers. India also leading the age standardized incidence of cervical cancer which is highest in South Asia at 22, compared to 19.2 in Bangladesh, 13 in Sri Lanka, and 2.8 in Ira
Data Loading...
