Scholarlink

<p>The rapid development of intelligent healthcare systems has greatly enhanced disease diagnosis using imaging technologies, but secure data management remains a challenge. Blockchain technology provides a natural cryptographic mechanism to ensure data integrity and protect against unauthorized access. Nevertheless, conventional medical imaging platforms are beset with security issues, computational inefficiency, and incorrect feature extraction. To resolve these challenges, a Blockchain-based Multi-Aspect Graph Attention Spherical Convolutional Neural Network with Augmented Snake Optimizer (MAGA-SPCNN-ASO) is introduced for disease diagnosis and the safe management of medical images. The approach starts with the acquisition of skin lesion images, and Dual Bilateral Least-Squares Hybrid Filtering (DBLSHF) is used to eliminate noise while preserving lesion details. Feature extraction is performed using the Multi-directional Shearlet Transform (MST), which efficiently extracts both fine and coarse features. The MAGA-SPCNN classifies the extracted features, and the cross-entropy loss function is trained using the Augmented Snake Optimizer (ASO). The processed images are securely encrypted using Chaos Fractal Digital Image Encryption (CFDIE) and stored using a Blockchain-based Convergence Media Consensus Mechanism (CMCM). The model achieves 99.3% accuracy, 99.1% precision, 99.4% recall, and 99.2% F1 score, with efficient execution time (30&#xa0;ms), low latency (70&#xa0;ms), and a high throughput rate (510 tps).</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

Blockchain-powered secure encryption for smart healthcare in industrial IoT with multi-aspect graph attention spherical convolutional neural network

  • 期刊论文

  • 发表日期：

    2026年1月10日

查看全文
  • Dipalee D. Rane Chaudhari,
  • D. Khalandar Basha,
  • Kingshuk Srivastava,
  • Mohd Naved,
  • Prolay Ghosh,
  • C. Rama Mohan

摘要

The rapid development of intelligent healthcare systems has greatly enhanced disease diagnosis using imaging technologies, but secure data management remains a challenge. Blockchain technology provides a natural cryptographic mechanism to ensure data integrity and protect against unauthorized access. Nevertheless, conventional medical imaging platforms are beset with security issues, computational inefficiency, and incorrect feature extraction. To resolve these challenges, a Blockchain-based Multi-Aspect Graph Attention Spherical Convolutional Neural Network with Augmented Snake Optimizer (MAGA-SPCNN-ASO) is introduced for disease diagnosis and the safe management of medical images. The approach starts with the acquisition of skin lesion images, and Dual Bilateral Least-Squares Hybrid Filtering (DBLSHF) is used to eliminate noise while preserving lesion details. Feature extraction is performed using the Multi-directional Shearlet Transform (MST), which efficiently extracts both fine and coarse features. The MAGA-SPCNN classifies the extracted features, and the cross-entropy loss function is trained using the Augmented Snake Optimizer (ASO). The processed images are securely encrypted using Chaos Fractal Digital Image Encryption (CFDIE) and stored using a Blockchain-based Convergence Media Consensus Mechanism (CMCM). The model achieves 99.3% accuracy, 99.1% precision, 99.4% recall, and 99.2% F1 score, with efficient execution time (30 ms), low latency (70 ms), and a high throughput rate (510 tps).