Breast Cancer Diagnosis Based on IoT and Deep Transfer Learning Enabled by Fog Computing.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Breast cancer poses a significant global health risk, with early diagnosis crucial for effective treatment.
• Technological advancements in medical image analysis and the Internet of Things (IoT) enable faster, more accurate disease detection.
• Early and remote diagnosis of chronic diseases like breast cancer is increasingly reliant on IoT integration.

Purpose of the Study:

• To develop an autonomous breast cancer diagnostic system using deep transfer learning (DTL) on mammography images.
• To enhance the accuracy and efficiency of breast cancer detection through advanced artificial intelligence techniques.
• To leverage DTL and Fog computing for secure, efficient, and high-performance medical image analysis.

Main Methods:

• Trained a DTL model using mammography images from The Cancer Imaging Archive (TCIA).
• Combined deep learning (DL) techniques like convolutional neural networks (CNNs) with transfer learning (TL) models (ResNet50, InceptionV3, AlexNet, VGG16, VGG19) and a support vector machine (SVM) classifier.
• Utilized Fog computing for enhanced data privacy, security, and reduced server load.

Main Results:

• The DTL model achieved high performance metrics: 97.99% accuracy, 99.51% precision, 98.43% sensitivity, 80.08% specificity, and 98.97% f1-score.
• The system demonstrated superior performance compared to existing methods on a large dataset of benign and malignant mammography images.
• Fog computing integration improved data security and system efficiency.

Conclusions:

• The proposed DTL model offers a viable and effective approach for autonomous breast cancer diagnosis.
• The system shows significant potential for improving early breast cancer detection rates and patient care.
• The integration of DTL and Fog computing presents a promising direction for secure and efficient medical diagnostic systems.