Systematic Evaluation of Vision Transformers for Automated Cervical Cancer Classification: Optimization, Statistical Validation, and Clinical Interpretability

TL;DR

This study systematically optimized Vision Transformer models for cervical cancer screening, achieving high accuracy and interpretability by aligning model attention with clinical features using the Herlev dataset.

Contribution

It introduces an optimized ViT architecture tailored for cervical cancer classification, demonstrating improved accuracy and interpretability over existing CNN-based methods.

Findings

Achieved 94.9%-95.2% cross-validation accuracy.

Identified effective augmentation and class weighting strategies.

Grad-CAM confirmed model attention aligns with clinical features.

Abstract

Manual Pap smear analysis for cervical cancer screening is limited by inter-observer variability, time constraints, and restricted expert availability. Although convolutional neural networks (CNNs) have automated cervical cell classification, they remain limited in modeling long-range spatial dependencies and often lack clinical interpretability. In this study, Vision Transformer (ViT) architectures were systematically optimized to enhance automated cervical cancer screening, which resulted in improved interpretability. The Herlev dataset (917 images: 242 normal, 675 abnormal) was utilized to optimize ViT-Tiny, a lightweight Vision Transformer architecture designed for reduced computational complexity, through a comprehensive evaluation of augmentation strategies, class weighting, and hyperparameters. The optimal configuration achieved 94.9%-95.2% cross-validation accuracy, in which random horizontal flipping and class weighting (0.7 x 1.3) were identified as most effective. Gradient-weighted Class Activation Mapping (Grad-CAM) analysis confirmed that model attention corresponded to clinically relevant morphological features, which include nuclear regions, cell boundaries, and chromatin texture, which align with cytopathological criteria. These findings indicate that Vision Transformers can deliver accurate and interpretable decision support for cervical cancer screening, which fulfills both clinical performance and transparency requirements essential for medical AI deployment.