Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Nov 1;20(11):3447-3456.
doi: 10.31557/APJCP.2019.20.11.3447.

Deep Convolution Neural Network for Malignancy Detection and Classification in Microscopic Uterine Cervix Cell Images

Shanthi P B  1 , Faraz Faruqi  1 , Hareesha K S  2 , Ranjini Kudva  3
Affiliations

Deep Convolution Neural Network for Malignancy Detection and Classification in Microscopic Uterine Cervix Cell Images

Shanthi P B et al. Asian Pac J Cancer Prev. .

Abstract

Objective: Automated Pap smear cervical screening is one of the most effective imaging based cancer detection<br /> tools used for categorizing cervical cell images as normal and abnormal. Traditional classification methods depend on<br /> hand-engineered features and show limitations in large, diverse datasets. Effective feature extraction requires an efficient<br /> image preprocessing and segmentation, which remains prominent challenge in the field of Pathology. In this paper, a<br /> deep learning concept is used for cell image classification in large datasets.

Methods: This relatively proposed novel<br /> method, combines abstract and complicated representations of data acquired in a hierarchical architecture. Convolution<br /> Neural Network (CNN) learns meaningful kernels that simulate the extraction of visual features such as edges, size,<br /> shape and colors in image classification. A deep prediction model is built using such a CNN network to classify the<br /> various grades of cancer: normal, mild, moderate, severe and carcinoma. It is an effective computational model which<br /> uses multiple processing layers to learn complex features. A large dataset is prepared for this study by systematically<br /> augmenting the images in Herlev dataset.

Result: Among the three sets considered for the study, the first set of single<br /> cell enhanced original images achieved an accuracy of 94.1% for 5 class, 96.2% for 4 class, 94.8% for 3 class and<br /> 95.7% for 2 class problems. The second set includes contour extracted images showed an accuracy of 92.14%, 92.9%,<br /> 94.7% and 89.9% for 5, 4, 3 and 2 class problems. The third set of binary images showed 85.07% for 5 class, 84%<br /> for 4 class, 92.07% for 3 class and highest accuracy of 99.97% for 2 class problems.

Conclusion: The experimental<br /> results of the proposed model showed an effective classification of different grades of cancer in cervical cell images,<br /> exhibiting the extensive potential of deep learning in Pap smear cell image classification.

Keywords: Cell image classification; Cervical screening; Convolution neural network; Deep Learning; Pap smear.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Enhanced Cervical Single Cell Images of Carcinoma, Severe, Moderate, Mild and Normal Grades Using Various Image Enhancement Techniques
Figure 2
Figure 2
Different Orientation of Augmented Images Generated From a Single Cervical Cell Image of Mild Grade
Figure 3
Figure 3
Architecture of Convolutional Neural Network Model Used for the Classification of Cervical Cell Images
Figure 4
Figure 4
Different Taxonomies of Cervical Cell Images Employed on 3 Different Sets for Cell Image Classification
Figure 5
Figure 5
Original Enhanced Single Cell Cervical Images of All 5 Grades (Set 1)
Figure 6
Figure 6
Contour Extracted Single Cell Cervical Images of All 5 Grades (Set 2)
Figure 7
Figure 7
Binarized Single Cell Nucleus Portion of Cervical Images of All 5 Grades (Set 3)
Figure 8
Figure 8
Comparison of Accuracies Obtained from the Proposed CNN Model for Different Classification Settings
Figure 9
Figure 9
Single Cell Pap Smear Cervical Cell Images of All 5 Grades Obtained from Kasturba Medical College, Manipal, India
Figure 10
Figure 10
Graph of Accuracy Vs. Percentage of Data Used while Training the Model and Indicates How Accuracy is Effected by the Size of the Training Set Size
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Anthimopoulos M, Christodoulidis S, Ebner L, et al. Lung pattern classification for interstitial lung diseases using a deep convolutional neural network. IEEE Trans Med Imaging. 2016;35:1207–16. - PubMed
    1. Arriaga-Garcia EF, Sanchez-Yanez RE, Garcia-Hernandez MG. Image enhancement using bi-histogram equalization with adaptive sigmoid functions in Proc Int. Conf IEEE CONIELECOMP. 2014:2261–4.
    1. Bora K, Chowdhury M, Mahanta LB, Kundu MK, Das AK. Pap smear image classification using convolutional neural network in Proc. 10th Ind Conf ACM ICVGIP. 2016: 55.
    1. Bruni L, Barrionuevo-Rosas L, Albero G, et al. Human Papillomavirus and Related Diseases in India. Summary Report. ICO Inf. Cent. HPV Cancer (HPV Inf. Centre) 2017:1–80.
    1. Devi MA, Ravi S, Vaishnavi J, Punitha S. Clas- sification of cervical cancer using artificial neural networks. Procedia Comput Sci. 2016;89:465–72.