Scholarlink

<p>Breast cancer (BC), a leading cause of women’s death in the world, can be greatly reduced with early and precise detection. Conventional detection methods are not generally very accurate or understandable. While deep learning models have shown promise, there are still challenges to precisely detecting complex patterns in medical images. To address these challenges, this work proposes a Hyperbolic Context Tangent Reverse Sigmoid Deep Attention Neural Networks with Dollmaker Optimization Algorithm (HCTRSDAN2Nets + DOA) model for BC detection and classification. The input images of the breasts are gathered from the CBIS-DDSM as well as BUSI databases and preprocessed initially utilizing the Sub-Aperture Keystone Transform Matched Filtering (SAKTMF) technique. Subsequently, Densenet201 with Single-Head Vision Transformer (Densenet201 + SHVT) is utilized for segmentation. For the detection and classification of BC, Hyperbolic Context Tangent Reverse Sigmoid Deep Attention Neural Networks (HCTRSDAN2Nets) are used, followed by the Dollmaker Optimization Algorithm (DOA) for optimization. With a 99.9% accuracy rate and 99.8% specificity, the CBIS-DDSM and BUSI datasets are employed to evaluate the effectiveness of the proposed technique, HCTRSDAN2Nets + DOA. The Python programming language is employed to achieve the proposed approach. The results of the suggested framework indicate excellent performance in the identification and classification of breast cancer with high precision, enhanced specificity, and uniform accuracy in a range of datasets. It successfully processes and splits up medical images with the strength of advanced deep neural networks and optimization techniques.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Breast Cancer Classification Using Optimized Hyperbolic Context Tangent Reverse Sigmoid Deep Attention Neural Networks

  • 期刊论文

  • 发表日期：

    2025年10月19日

查看全文
  • Huma Khan,
  • Dinesh Kumar Anguraj,
  • D. Pavithra,
  • Krishna Prakash Arunachalam

摘要

Breast cancer (BC), a leading cause of women’s death in the world, can be greatly reduced with early and precise detection. Conventional detection methods are not generally very accurate or understandable. While deep learning models have shown promise, there are still challenges to precisely detecting complex patterns in medical images. To address these challenges, this work proposes a Hyperbolic Context Tangent Reverse Sigmoid Deep Attention Neural Networks with Dollmaker Optimization Algorithm (HCTRSDAN2Nets + DOA) model for BC detection and classification. The input images of the breasts are gathered from the CBIS-DDSM as well as BUSI databases and preprocessed initially utilizing the Sub-Aperture Keystone Transform Matched Filtering (SAKTMF) technique. Subsequently, Densenet201 with Single-Head Vision Transformer (Densenet201 + SHVT) is utilized for segmentation. For the detection and classification of BC, Hyperbolic Context Tangent Reverse Sigmoid Deep Attention Neural Networks (HCTRSDAN2Nets) are used, followed by the Dollmaker Optimization Algorithm (DOA) for optimization. With a 99.9% accuracy rate and 99.8% specificity, the CBIS-DDSM and BUSI datasets are employed to evaluate the effectiveness of the proposed technique, HCTRSDAN2Nets + DOA. The Python programming language is employed to achieve the proposed approach. The results of the suggested framework indicate excellent performance in the identification and classification of breast cancer with high precision, enhanced specificity, and uniform accuracy in a range of datasets. It successfully processes and splits up medical images with the strength of advanced deep neural networks and optimization techniques.