Automated Quality Evaluation of Cervical Cytopathology Whole Slide Images Based on Content Analysis

TL;DR

This work introduces a fully automated, content-aware quality assessment framework for cervical cytopathology WSIs that aligns with The Bethesda System (TBS). It combines multiple deep learning models—grid-artifact analysis, FocusAttNet for focus, DoubleUNet for artifact segmentation, staining standard estimation, and YOLOv5 for cellular content—into a unified metric system, and uses XGBoost to fuse patch-level metrics into an overall WSI score. The approach achieves faster, more consistent evaluations than manual methods, validated on 302 WSIs with 33k+ training patches and pathologist-rated quality scores, and demonstrates effective separation of quality issues stemming from preparation or scanning. This method enables scalable QC for CAD pipelines and supports decision-making for slide re-preparation or re-scanning in clinical workflows, with potential generalization to other cytopathology tasks.

Abstract

The ThinPrep Cytologic Test (TCT) is the most widely used method for cervical cancer screening, and the sample quality directly impacts the accuracy of the diagnosis. Traditional manual evaluation methods rely on the observation of pathologist under microscopes. These methods exhibit high subjectivity, high cost, long duration, and low reliability. With the development of computer-aided diagnosis (CAD), an automated quality assessment system that performs at the level of a professional pathologist is necessary. To address this need, we propose a fully automated quality assessment method for Cervical Cytopathology Whole Slide Images (WSIs) based on The Bethesda System (TBS) diagnostic standards, artificial intelligence algorithms, and the characteristics of clinical data. The method analysis the context of WSIs to quantify quality evaluation metrics which are focused by TBS such as staining quality, cell counts and cell mass proportion through multiple models including object detection, classification and segmentation. Subsequently, the XGBoost model is used to mine the attention paid by pathologists to different quality evaluation metrics when evaluating samples, thereby obtaining a comprehensive WSI sample score calculation model. Experimental results on 100 WSIs demonstrate that the proposed evaluation method has significant advantages in terms of speed and consistency.

Figures (10)

• Figure 1: Pipeline of the proposed method.
• Figure 2: Model for Segmenting Marker and Air/gel Bubbles.
• Figure 3: Cervical Cell Staining Separation Process.
• Figure 4: Network structure of YOLOV5.
• Figure 5: Examples of quality issues in cervical cytopathology images.