Robust whole slide image analysis for cervical cancer screening using deep learning

TLDR

Computer‑assisted diagnosis is key for scaling up cervical cancer screening, yet current recognition algorithms perform poorly on whole slide images, fail to generalize across diverse staining and imaging, and show sub‑optimal clinical verification. We aim to develop a progressive lesion cell recognition method that combines low‑ and high‑resolution WSIs to recommend lesion cells and a recurrent neural network‑based WSI classification model to evaluate lesion degree. The system is trained and validated on 3,545 patient‑wise WSIs with 79,911 annotations from multiple hospitals and imaging instruments, and uses a recurrent neural network for WSI classification. On multi‑center independent test sets of 1,170 patient‑wise WSIs, the system achieves 93.5 % specificity and 95.1 % sensitivity for slide classification, outperforming the average performance of three cytopathologists, and attains an 88.5 % true‑positive rate for highlighting the top 10 lesion cells on 447 positive slides, with a processing time of about 1.5 min per giga‑pixel WSI.

Abstract

Computer-assisted diagnosis is key for scaling up cervical cancer screening. However, current recognition algorithms perform poorly on whole slide image (WSI) analysis, fail to generalize for diverse staining and imaging, and show sub-optimal clinical-level verification. Here, we develop a progressive lesion cell recognition method combining low- and high-resolution WSIs to recommend lesion cells and a recurrent neural network-based WSI classification model to evaluate the lesion degree of WSIs. We train and validate our WSI analysis system on 3,545 patient-wise WSIs with 79,911 annotations from multiple hospitals and several imaging instruments. On multi-center independent test sets of 1,170 patient-wise WSIs, we achieve 93.5% Specificity and 95.1% Sensitivity for classifying slides, comparing favourably to the average performance of three independent cytopathologists, and obtain 88.5% true positive rate for highlighting the top 10 lesion cells on 447 positive slides. After deployment, our system recognizes a one giga-pixel WSI in about 1.5 min.