Benchmarking Swin Transformers for Histopathological Image-Based Breast Cancer Detection

Breast cancer remains a leading cause of mortality among women, where early and accurate diagnosis is critical for effective treatment. However, classifying histopathological images remains challenging due to variability in tissue morphology and image quality. This study uses the BreaKHis dataset, enhanced with Contrast Limited Adaptive Histogram Equalization (CLAHE), to address binary classification (benign vs. malignant) across four magnification levels $(40 \mathrm{X}, 100 \mathrm{X}, 200 \mathrm{X}$, 400X). A comparative analysis of Swin Transformer V2 (Swin V2) variants (Tiny, Small, and Base) is conducted to evaluate performance using accuracy, precision, recall, F1-score, and Area Under the Curve (AUC). The Swin V2 Small variant achieves the highest accuracy of 99.45 % and AUC of 99.95 %. Future research will concentrate on testing the model using data from multiple institutions and adapting it for efficient use in clinical environments with limited resources.

• Publication year

  2025

• Venue

  2025 International Conference on Emerging Trends in Networks and Computer Communications (ETNCC)

• Publication date

  2025-08-05

• Fields of study

  Not labeled

• Identifiers
  DOI 10.1109/ETNCC66224.2025.11299684
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Advancing breast cancer diagnosis: token vision transformers for faster and accurate classification of histopathology images

  2025influential reference
• Vision Transformers and CNN-Based Knowledge-Distillation for Histopathological Image Classification

  2024influential reference
• Attention-Based Deep Learning Approach for Breast Cancer Histopathological Image Multi-Classification

  2024influential reference
• High-Performance Classification of Breast Cancer Histopathological Images Using Fine-Tuned Vision Transformers on the BreakHis Dataset

  2024influential reference
• ST-Double-Net: A Two-Stage Breast Tumor Classification Model Based on Swin Transformer and Weakly Supervised Target Localization

  2024influential reference
• Histopathological Breast Cancer Detection: A Vision Transformer and Attention-Based Approach

  2024cited by this paper
• Breast Cancer Classification Using Fine-Tuned SWIN Transformer Model on Mammographic Images

  2024influential reference
• SwinCNN: An Integrated Swin Transformer and CNN for Improved Breast Cancer Grade Classification

  2024influential reference
• Semi-supervised vision transformer with adaptive token sampling for breast cancer classification

  2022influential reference
• ViT-DeiT: An Ensemble Model for Breast Cancer Histopathological Images Classification

  2022cited by this paper
• Swin Transformer V2: Scaling Up Capacity and Resolution

  2021influential reference
• A Dataset for Breast Cancer Histopathological Image Classification

  2016cited by this paper
• Contrast Limited Adaptive Histogram Equalization

  1994cited by this paper

• No citing papers are available for this paper.