Artificial Intelligence in Cervical Cancer Screening and Diagnosis

Abstract

Cervical cancer remains a leading cause of cancer death in women, seriously threatening their physical and mental health. It is an easily preventable cancer with early screening and diagnosis. Although technical advancements have significantly improved the early diagnosis of cervical cancer, accurate diagnosis remains difficult owing to various factors. In recent years, artificial intelligence (AI)-based medical diagnostic applications have been on the rise and have excellent applicability in the screening and diagnosis of cervical cancer. Their benefits include reduced time consumption, reduced need for professional and technical personnel, and no bias owing to subjective factors. We, thus, aimed to discuss how AI can be used in cervical cancer screening and diagnosis, particularly to improve the accuracy of early diagnosis. The application and challenges of using AI in the diagnosis and treatment of cervical cancer are also discussed.

Keywords: artificial intelligence; cervical cancer; cervical intraepithelial neoplasia (CIN); colposcopy; cytology; deep learning; early screening and diagnosis.

Figure 1

Evolution of cervical cancer screening methods. The figure shows major milestones in the evolution of cervical cancer screening. The main screening methods for cervical cancer are HPV testing and TCT (cytology) nowadays.

Figure 2

Cervical cancer screening procedures are recommended for women aged 25 to 65. The American Cancer Society recommends screening starting at age 25 Colposcopy is recommended for HPV16/18 +, ASC-US and high risk HPV+ cytology with cytological results above ASC-H. Re-screening is recommended after 1 year for other abnormalities, and after 3 years for normal ones.

Figure 3

A example segmentation model based on Mask-RCNN architecture. Reproduced with the permission of ref. (51), copyright@IEEE, 2019. In training phase, input was pap smear slide image and nucleus ground truth mask with class label was preprocessed and then trained in Mask R-CNN. In testing phase, pap smear slide image was preprocessed. Mask RCNN was used to specify bounding box, nucleus mask, and class of each cell.

Figure 4

Schematic representation of application of Convolutional Neural Network in colposcopy images. Schematic depicting that a CNN pre-trained on other large-scale image datasets can be adapted to significantly increase the accuracy and shorten the training duration of a network trained on a novel dataset of colposcopy images.