Scholarlink

<p>This paper addresses critical security and accuracy challenges in artificial intelligence (AI)-driven remote healthcare monitoring by proposing an integrated framework combining three components: Firstly, the ‘Lasso Regression Ranking with Modswish based Deep Learning Neural Network (LRRM-DLNN)’ algorithm for simultaneous disease classification which achieved 98.9% accuracy and malicious activity prediction with 99% accuracy. Secondly, the ‘Centre Point Pi-membership Functional Fuzzy Rule (CPPFFR)’ system for AI data preservation with a 654–682&#xa0;ms processing speed. Thirdly, Elliptic Square Root Curve Cryptography (ESRCC) encryption uses optimised curve computation with 1093&#xa0;ms/1142&#xa0;ms processing speed for encryption/decryption. Implemented on Python/SQL with a 16-core CPU/32&#xa0;GB RAM/NVIDIA GPU infrastructure and Google Colab, the system integrates cryptographic puzzle (CP) authentication and Gaussian Functional Hippopotamus Optimisation (GFHO)-based feature selection, reducing time complexity by 38%. Experimental validation using NSL-KDD and clinical datasets demonstrates superiority over existing models, achieving 96.7% F-measure in threat detection, 96.6% in disease diagnosis, and around 4.8 RMSE error. Dual smart contracts resolve 92% of blockchain errors without reconstruction. The framework outperforms Bi-LSTM and CNN baselines by 19.8% accuracy and 45% faster threat detection, establishing a new benchmark for secure, efficient healthcare monitoring in resource-constrained environments.</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

Improved Health Care Monitoring System Using Proposed LRRN-DLNN Algorithm Based on CPPFFR AI Data Preservation and ESRCC-Based Security

  • 期刊论文

  • 发表日期：

    2025年7月14日

查看全文
  • Yashwant Aditya,
  • Priyank Jain,
  • Ritu Tiwari

摘要

This paper addresses critical security and accuracy challenges in artificial intelligence (AI)-driven remote healthcare monitoring by proposing an integrated framework combining three components: Firstly, the ‘Lasso Regression Ranking with Modswish based Deep Learning Neural Network (LRRM-DLNN)’ algorithm for simultaneous disease classification which achieved 98.9% accuracy and malicious activity prediction with 99% accuracy. Secondly, the ‘Centre Point Pi-membership Functional Fuzzy Rule (CPPFFR)’ system for AI data preservation with a 654–682 ms processing speed. Thirdly, Elliptic Square Root Curve Cryptography (ESRCC) encryption uses optimised curve computation with 1093 ms/1142 ms processing speed for encryption/decryption. Implemented on Python/SQL with a 16-core CPU/32 GB RAM/NVIDIA GPU infrastructure and Google Colab, the system integrates cryptographic puzzle (CP) authentication and Gaussian Functional Hippopotamus Optimisation (GFHO)-based feature selection, reducing time complexity by 38%. Experimental validation using NSL-KDD and clinical datasets demonstrates superiority over existing models, achieving 96.7% F-measure in threat detection, 96.6% in disease diagnosis, and around 4.8 RMSE error. Dual smart contracts resolve 92% of blockchain errors without reconstruction. The framework outperforms Bi-LSTM and CNN baselines by 19.8% accuracy and 45% faster threat detection, establishing a new benchmark for secure, efficient healthcare monitoring in resource-constrained environments.