A Lightweight, Interpretable Deep Learning System for Automated Detection of Cervical Adenocarcinoma In Situ (AIS)

TL;DR

This paper presents a lightweight, interpretable deep learning system that accurately detects cervical adenocarcinoma in situ from histology images, aiding early diagnosis and potentially improving screening and educational efforts.

Contribution

The study introduces a novel, efficient deep learning model with interpretability features for AIS detection, trained on a large, expert-labeled dataset, and deployed as a virtual diagnostic tool.

Findings

Achieved 73.23% accuracy in AIS detection

Demonstrated biologically interpretable Grad-CAM heatmaps

Potential for use in screening and low-resource settings

Abstract

Cervical adenocarcinoma in situ (AIS) is a critical premalignant lesion whose accurate histopathological diagnosis is challenging. Early detection is essential to prevent progression to invasive cervical adenocarcinoma. In this study, we developed a deep learning-based virtual pathology assistant capable of distinguishing AIS from normal cervical gland histology using the CAISHI dataset, which contains 2240 expert-labeled H&E images (1010 normal and 1230 AIS). All images underwent Macenko stain normalization and patch-based preprocessing to enhance morphological feature representation. An EfficientNet-B3 convolutional neural network was trained using class-balanced sampling and focal loss to address dataset imbalance and emphasize difficult examples. The final model achieved an overall accuracy of 0.7323, with an F1-score of 0.75 for the Abnormal class and 0.71 for the Normal class. Grad-CAM heatmaps demonstrated biologically interpretable activation patterns, highlighting nuclear atypia and glandular crowding consistent with AIS morphology. The trained model was deployed in a Gradio-based virtual diagnostic assistant. These findings demonstrate the feasibility of lightweight, interpretable AI systems for cervical gland pathology, with potential applications in screening workflows, education, and low-resource settings.