Scholarlink

This study examines various machine learning (ML) and deep learning (DL) algorithms for the detection of breast cancer using the Mammographic Image Analysis Society (MIAS) dataset. The MIAS dataset has been augmented using data augmentation techniques to expand the total to approximately 7,000 images. Among the models examined in this study were convolutional neural networks (CNN), support vector machines (SVM), random forests, artificial neural networks (ANN), long short-term memory (LSTM), VGG-16, AlexNet, and stochastic gradient descent (SGD). Python was utilized for feature extraction and data preprocessing, with augmentation methods applied to enhance the model’s robustness. Performance was evaluated using F1-score, AUC-ROC, recall, accuracy, and precision. Both the Random Forest and CNN models demonstrated potential for application in clinical settings, achieving the highest classification accuracies of 98.45% and 96.99%, respectively. The study also addresses the model’s limitations, generalizability issues, and future research opportunities, including model interpretability and multi-modal data integration. The findings indicate that using AI techniques can significantly improve early detection of breast cancer and support medical decision-making.

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

Comparative Study of Machine and Deep Learning Models for Breast Cancer Detection Using Medical Imaging Data

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Lalah Khadeejah Mustafa Al Sharif,
  • Ibrahim Soliman Al-Saedi,
  • Asma Mustafa Al Sharif,
  • Walid T. Shanab

摘要

This study examines various machine learning (ML) and deep learning (DL) algorithms for the detection of breast cancer using the Mammographic Image Analysis Society (MIAS) dataset. The MIAS dataset has been augmented using data augmentation techniques to expand the total to approximately 7,000 images. Among the models examined in this study were convolutional neural networks (CNN), support vector machines (SVM), random forests, artificial neural networks (ANN), long short-term memory (LSTM), VGG-16, AlexNet, and stochastic gradient descent (SGD). Python was utilized for feature extraction and data preprocessing, with augmentation methods applied to enhance the model’s robustness. Performance was evaluated using F1-score, AUC-ROC, recall, accuracy, and precision. Both the Random Forest and CNN models demonstrated potential for application in clinical settings, achieving the highest classification accuracies of 98.45% and 96.99%, respectively. The study also addresses the model’s limitations, generalizability issues, and future research opportunities, including model interpretability and multi-modal data integration. The findings indicate that using AI techniques can significantly improve early detection of breast cancer and support medical decision-making.