Abstract
Cervical cancer is a worldwide public health problem with a high rate of illness and mortality among women. In this study, we proposed a novel framework based on Faster RCNN-FPN architecture for the detection of abnormal cervical cells in cytology images from a cancer screening test. We extended the Faster RCNN-FPN model by infusing deformable convolution layers into the feature pyramid network (FPN) to improve scalability. Furthermore, we introduced a global contextual aware module alongside the Region Proposal Network (RPN) to enhance the spatial correlation between the background and the foreground. Extensive experimentations with the proposed deformable and global context aware (DGCA) RCNN were carried out using the cervical image dataset of “Digital Human Body” Vision Challenge from the Alibaba Cloud TianChi Company. Performance evaluation based on the mean average precision (mAP) and receiver operating characteristic (ROC) curve has demonstrated considerable advantages of the proposed framework. Particularly, when combined with tagging of the negative image samples using traditional computer-vision techniques, 6–9% increase in mAP has been achieved. The proposed DGCA-RCNN model has potential to become a clinically useful AI tool for automated detection of cervical cancer cells in whole slide images of Pap smear.
Highlights
Cervical cancer is the second most common malignancy among women with more than half a million new cases reported annually
Relevant concerns include the following: (1) the abnormalities in the cancerous cells are subtle and complex; (2) Cells in the cervical Pap smear images exist in different sizes and their geometric shapes can be obscured due to clump formation and overlapping with artifacts; (3) The region of interest (ROI) pooling/aligning processes in the object detection algorithms tend to enhance the local receptive fields and lose global context information. Aimed at mitigating these potential limiting factors for more accurate detection of abnormal cells in cervical cytology images, in this study, we introduced two functional extensions for one of the top-performing object detectors, the Faster RCNN-feature pyramid network (FPN) framework: One is the infusion of deformable convolution network (DCN) [47,48,49,50,51] in the last three stages of the bottom-up pathway of the FPN and the other is adding a global context aware (GCA) module [52] alongside the region proposal network (RPN)
We evaluated the performance of the proposed deformable and global context aware (DGCA)-RCNN framework and conducted systematic performance comparison with five other state-of-the-art object detectors including ATSS [53], RetinaNet [39], Faster RCNN-FPN [34], double-head [54], and Cascade RCNN [32]
Summary
Cervical cancer is the second most common malignancy among women with more than half a million new cases reported annually. It can be detected and prevented at its early stage by cytology screening test. Cytological slides of vaginal smear of Pap strains are microscopically examined at 400× magnification. Even with experienced pathologists, it is a tedious task to examine the Pap smear slides through microscope and the detection of cervical cancer cells can often be missed, due to the small size of the cervical intraepithelial neoplasia, overlapping clump of cells or masking by blood mucus and artifacts
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have