Scholarlink

<p>Breast cancer is the second most common cause of cancer death in women and one of the worst diseases. Breast cancer arises when breast cells start to grow into malignant, cancerous tumours. Regular professional exams and self-tests help with early diagnosis, which improves survival chances. The classification of breast cancer as a medical procedure is a significant challenge for researchers. The primary goal of this project is to use 3D pictures and AI monitoring to detect breast cancer earlier. To enhance image quality, the mammography pictures are first gathered and pre-processed. This research proposes novel methods in Validating Multimodal Breast Cancer Validation with Numerical and 3D Image Datasets using Hybrid Transfer Learning methods. Here input is collected as numerical data with 3D breast image and processed for noise removal as well as normalization. Then this image features have been extracted as well as classified utilizing convolutional graph adversarial gaussian neural network (ConGAdGNN) and the extracted feature classification is carried out using hybrid transfer attention auto-encoder BiLSTM (HTrAAEBiLSTM) model. The experimental analysis is carried out in terms of Training accuracy, Validation accuracy, Average precision, RRMSE, Recall, F-measure. Proposed technique Training accuracy of 97%, Validation accuracy of 98%, RECALL of 93%, Average precision of 95%, RMSE of 60%.</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

3D Image Based Breast Cancer Validation by Classification and Feature Extraction Using Hybrid Transfer Learning Techniques

  • 期刊论文

  • 发表日期：

    2026年1月14日

查看全文
  • Syambabu Badugu,
  • V. D. Ambeth Kumar,
  • S. Hemamalini,
  • V. Sathya Preiya

摘要

Breast cancer is the second most common cause of cancer death in women and one of the worst diseases. Breast cancer arises when breast cells start to grow into malignant, cancerous tumours. Regular professional exams and self-tests help with early diagnosis, which improves survival chances. The classification of breast cancer as a medical procedure is a significant challenge for researchers. The primary goal of this project is to use 3D pictures and AI monitoring to detect breast cancer earlier. To enhance image quality, the mammography pictures are first gathered and pre-processed. This research proposes novel methods in Validating Multimodal Breast Cancer Validation with Numerical and 3D Image Datasets using Hybrid Transfer Learning methods. Here input is collected as numerical data with 3D breast image and processed for noise removal as well as normalization. Then this image features have been extracted as well as classified utilizing convolutional graph adversarial gaussian neural network (ConGAdGNN) and the extracted feature classification is carried out using hybrid transfer attention auto-encoder BiLSTM (HTrAAEBiLSTM) model. The experimental analysis is carried out in terms of Training accuracy, Validation accuracy, Average precision, RRMSE, Recall, F-measure. Proposed technique Training accuracy of 97%, Validation accuracy of 98%, RECALL of 93%, Average precision of 95%, RMSE of 60%.