Automated Cervical Digitized Histology Whole-Slide Image Analysis Toolbox

Abstract

Background: Cervical intraepithelial neoplasia (CIN) is regarded as a potential precancerous state of the uterine cervix. Timely and appropriate early treatment of CIN can help reduce cervical cancer mortality. Accurate estimation of CIN grade correlated with human papillomavirus type, which is the primary cause of the disease, helps determine the patient's risk for developing the disease. Colposcopy is used to select women for biopsy. Expert pathologists examine the biopsied cervical epithelial tissue under a microscope. The examination can take a long time and is prone to error and often results in high inter-and intra-observer variability in outcomes. Methodology: We propose a novel image analysis toolbox that can automate CIN diagnosis using whole slide image (digitized biopsies) of cervical tissue samples. The toolbox is built as a four-step deep learning model that detects the epithelium regions, segments the detected epithelial portions, analyzes local vertical segment regions, and finally classifies each epithelium block with localized attention. We propose an epithelium detection network in this study and make use of our earlier research on epithelium segmentation and CIN classification to complete the design of the end-to-end CIN diagnosis toolbox. Results: The results show that automated epithelium detection and segmentation for CIN classification yields comparable results to manually segmented epithelium CIN classification. Conclusion: This highlights the potential as a tool for automated digitized histology slide image analysis to assist expert pathologists.

An Integrated Approach to Automated Diagnosis of Cervical Intraepithelial Neoplasia in Digital Histology Images

“…Repeated refinement of the data, and training parameters, based on evaluation results, enabled achievement of 92.14% accuracy for the model [5], further selected as a benchmark with mean F-score (92.19%), indicating excellent agreement between the model's predictions and the true CIN values. Mean kappa (89.4%) and weighted kappa (90.83%) also indicate good agreement.…”

“…The cervical WSI image SSE extraction model [5] adopted for multiclass classification of histopathology images and combined with A2 fusion demonstrated the best performance to detect CIN 1 and CIN 2 in this dataset. However, the fusion method comparison was limited because only a few samples had conflicting predictions within their patch group.…”

Cervical Intraepithelial Neoplasia Grading from Prepared Digital Histology Images

“…Repeated refinement of the data, and training parameters, enabled achievement of 92.14% accuracy for CNN model [5], adapted for multiclass classification, further selected as a benchmark (Fig. 2).…”

“…Moreover, histopathological images themselves pose some unique challenges. Image appearance is highly variable due to slide preparation, the nonstandard shape of the biopsied tissue and size of the epithelium regions, as well as the presence of artefacts, e.g., stains, ink markers, tapes, and blurred regions [4,5]. Hence, wide adoption of deep learning models in digital pathology is limited.…”