论文信息 - A comparative study of 3 deep learning models for Pap smear screening

A comparative study of 3 deep learning models for Pap smear screening

This work presents a comparative study of automated screening procedure for Pap smear with deep learning technology. Three convolution neural network models (AlexNet, densenet161 and resnet101) were employed for detecting the presence of cervical precancerous or cancerous cells from Pap smear database. The study compares accuracy, sensitivity, specificity, and computation time for each deep learning model. The best model is the densenet161 due to its high sensitivity and accuracy which are key factors in an automated Pap smear screening procedure to offer the best early detection of cervical cancer to better treatment outcomes.

C. Pintavirooj | Wibool Piyawattanametha | Yuttachon Promworn

[1] Michael Quinn,et al. Chapter 8: Screening for cervical cancer in developing countries. , 2006, Vaccine.

[2] Jian Sun,et al. Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3] S. Fedewa,et al. Cancer screening in the United States, 2015: A review of current American Cancer Society guidelines and current issues in cancer screening , 2015, CA: a cancer journal for clinicians.

[4] Alex Krizhevsky,et al. One weird trick for parallelizing convolutional neural networks , 2014, ArXiv.

[5] Gloria Bueno,et al. Glomerulus Classification with Convolutional Neural Networks , 2017, MIUA.