Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning

Accurate and early breast cancer detection is critical for improving patient outcomes. In this study, we propose PatchCascade-ViT, a novel self-supervised Vision Transformer (ViT) framework for automated BI-RADS classification of mammographic images. Unlike conventional deep learning approaches that...

Full description

Saved in:
Bibliographic Details
Main Authors: Abdelrahman Abdallah, Mahmoud Salaheldin Kasem, Ibrahim Abdelhalim, Norah Saleh Alghamdi, Ayman El-Baz
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11045361/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849391679591677952
author Abdelrahman Abdallah
Mahmoud Salaheldin Kasem
Ibrahim Abdelhalim
Norah Saleh Alghamdi
Ayman El-Baz
author_facet Abdelrahman Abdallah
Mahmoud Salaheldin Kasem
Ibrahim Abdelhalim
Norah Saleh Alghamdi
Ayman El-Baz
author_sort Abdelrahman Abdallah
collection DOAJ
description Accurate and early breast cancer detection is critical for improving patient outcomes. In this study, we propose PatchCascade-ViT, a novel self-supervised Vision Transformer (ViT) framework for automated BI-RADS classification of mammographic images. Unlike conventional deep learning approaches that rely heavily on annotated datasets, PatchCascade-ViT leverages Self Patch-level Supervision (SPS) to learn meaningful mammographic representations from unlabeled data, significantly enhancing classification performance. Our framework operates through a two-stage cascade classification process. In the first stage, the model differentiates non-cancerous from potentially cancerous mammograms using SelfPatch, an innovative self-supervised learning task that enhances patch-level feature learning by enforcing consistency among spatially correlated patches. The second stage refines the classification by distinguishing Scattered Fibroglandular from Heterogeneously and Extremely Dense breast tissue categories, enabling more precise breast cancer risk assessment. To validate the effectiveness of PatchCascade-ViT, we conducted extensive evaluations on a dataset of 4,368 mammograms across three BI-RADS classes. Our method achieved a system sensitivity of 85.01% and an F1-score of 84.90%, outperforming existing deep learning-based approaches. By integrating self-supervised learning with a cascade vision transformer architecture, PatchCascade-ViT reduces reliance on annotated datasets while maintaining high classification accuracy. These findings demonstrate its potential for enhancing breast cancer screening, aiding radiologists in early detection, and improving clinical decision-making.
format Article
id doaj-art-7aadcf898b054c4b9d0e66c5fbfe552c
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-7aadcf898b054c4b9d0e66c5fbfe552c2025-08-20T03:40:59ZengIEEEIEEE Access2169-35362025-01-011313550013551410.1109/ACCESS.2025.358158211045361Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade LearningAbdelrahman Abdallah0https://orcid.org/0000-0001-8747-4927Mahmoud Salaheldin Kasem1https://orcid.org/0000-0002-8513-570XIbrahim Abdelhalim2https://orcid.org/0009-0000-1544-7276Norah Saleh Alghamdi3https://orcid.org/0000-0001-6421-6001Ayman El-Baz4https://orcid.org/0000-0001-7264-1323Department of Information Technology, Assiut University, Assiut, EgyptDepartment of Computer Science, University of Innsbruck, Innsbruck, AustriaDepartment of Bioengineering, University of Louisville, Louisville, KY, USADepartment of Computer Sciences, Princess Nourah bint Abdulrahman University, Riyadh, Saudi ArabiaDepartment of Bioengineering, University of Louisville, Louisville, KY, USAAccurate and early breast cancer detection is critical for improving patient outcomes. In this study, we propose PatchCascade-ViT, a novel self-supervised Vision Transformer (ViT) framework for automated BI-RADS classification of mammographic images. Unlike conventional deep learning approaches that rely heavily on annotated datasets, PatchCascade-ViT leverages Self Patch-level Supervision (SPS) to learn meaningful mammographic representations from unlabeled data, significantly enhancing classification performance. Our framework operates through a two-stage cascade classification process. In the first stage, the model differentiates non-cancerous from potentially cancerous mammograms using SelfPatch, an innovative self-supervised learning task that enhances patch-level feature learning by enforcing consistency among spatially correlated patches. The second stage refines the classification by distinguishing Scattered Fibroglandular from Heterogeneously and Extremely Dense breast tissue categories, enabling more precise breast cancer risk assessment. To validate the effectiveness of PatchCascade-ViT, we conducted extensive evaluations on a dataset of 4,368 mammograms across three BI-RADS classes. Our method achieved a system sensitivity of 85.01% and an F1-score of 84.90%, outperforming existing deep learning-based approaches. By integrating self-supervised learning with a cascade vision transformer architecture, PatchCascade-ViT reduces reliance on annotated datasets while maintaining high classification accuracy. These findings demonstrate its potential for enhancing breast cancer screening, aiding radiologists in early detection, and improving clinical decision-making.https://ieeexplore.ieee.org/document/11045361/Breast cancerBI-RADS classificationself-supervised learning (SSL)vision transformers (ViTs)cascade classification
spellingShingle Abdelrahman Abdallah
Mahmoud Salaheldin Kasem
Ibrahim Abdelhalim
Norah Saleh Alghamdi
Ayman El-Baz
Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
IEEE Access
Breast cancer
BI-RADS classification
self-supervised learning (SSL)
vision transformers (ViTs)
cascade classification
title Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
title_full Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
title_fullStr Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
title_full_unstemmed Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
title_short Improving BI-RADS Mammographic Classification With Self-Supervised Vision Transformers and Cascade Learning
title_sort improving bi rads mammographic classification with self supervised vision transformers and cascade learning
topic Breast cancer
BI-RADS classification
self-supervised learning (SSL)
vision transformers (ViTs)
cascade classification
url https://ieeexplore.ieee.org/document/11045361/
work_keys_str_mv AT abdelrahmanabdallah improvingbiradsmammographicclassificationwithselfsupervisedvisiontransformersandcascadelearning
AT mahmoudsalaheldinkasem improvingbiradsmammographicclassificationwithselfsupervisedvisiontransformersandcascadelearning
AT ibrahimabdelhalim improvingbiradsmammographicclassificationwithselfsupervisedvisiontransformersandcascadelearning
AT norahsalehalghamdi improvingbiradsmammographicclassificationwithselfsupervisedvisiontransformersandcascadelearning
AT aymanelbaz improvingbiradsmammographicclassificationwithselfsupervisedvisiontransformersandcascadelearning