Towards Improved Cervical Cancer Screening: Vision Transformer-Based Classification and Interpretability

ABSTRACT We propose a novel approach to cervical cell image classification for cervical cancer screening using the EV A-02 transformer model. We developed a four-step pipeline: fine-tuning EV A-02, feature extraction, selecting important features through multiple machine learning models, and training a new artificial neural network with optional loss weighting for improved generalization. With this design, our best model achieved an F1-score of 0.85227, outperforming the baseline EV A-02 model (0.84878). We also utilized Kernel SHAP analysis and identified key features correlating with cell morphology and staining characteristics, providing interpretable insights into the decision-making process of the fine-tuned model. Index T erms-- Cell Classification, Cervical Cancer, Explainable AI, Vision Transformers 1. INTRODUCTION Cervical cancer remains a significant global health challenge, ranking as the fourth most common cancer among women with over 600,000 new cases and 300,000 deaths annually [1].