Rafiq Adnan, Jaffar Arfan, Latif Ghazanfar, Masood Sohail, Abdelhamid Sherif E
Department of Computer Science & IT, Superior University, Lahore 54000, Pakistan.
Department of Computing Science, Thompson Rivers University, Kamloops, BC V2C 0C8, Canada.
Diagnostics (Basel). 2025 Feb 27;15(5):582. doi: 10.3390/diagnostics15050582.
Breast cancer is among the most frequently diagnosed cancers and leading cause of mortality worldwide. The accurate classification of breast cancer from the histology photographs is very important for the diagnosis and effective treatment planning. : In this article, we propose a DenseNet121-based deep learning model for breast cancer detection and multi-class classification. The experiments were performed using whole-slide histopathology images collected from the BreakHis dataset. : The proposed method attained state-of-the-art performance with a 98.50% accuracy and an AUC of 0.98 for the binary classification. In multi-class classification, it obtained competitive results with 92.50% accuracy and an AUC of 0.94. : The proposed model outperforms state-of-the-art methods in distinguishing between benign and malignant tumors as well as in classifying specific malignancy subtypes. This study highlights the potential of deep learning in breast cancer diagnosis and establishes the foundation for developing advanced diagnostic tools.
乳腺癌是全球最常被诊断出的癌症之一，也是主要的死亡原因。从组织学照片中准确分类乳腺癌对于诊断和有效的治疗规划非常重要。在本文中，我们提出了一种基于DenseNet121的深度学习模型用于乳腺癌检测和多类别分类。实验使用了从BreakHis数据集中收集的全切片组织病理学图像。所提出的方法在二分类中达到了98.50%的准确率和0.98的AUC，取得了当前最优的性能。在多类别分类中，它获得了具有竞争力的结果，准确率为92.50%，AUC为0.94。所提出的模型在区分良性和恶性肿瘤以及对特定恶性亚型进行分类方面优于当前最优方法。这项研究突出了深度学习在乳腺癌诊断中的潜力，并为开发先进的诊断工具奠定了基础。
