Related Experiment Video
Updated: Jun 17, 2026
06:32
Bringing the Clinic Home: An At-Home Multi-Modal Data Collection Ecosystem to Support Adaptive Deep Brain Stimulation
Published on: July 14, 2023
Remote medical system driven by medical big models: Dynamic defense model for network security threats
Zhenfeng Weng1,2, Yining Hu1, Dayong Gu2
1School of Cyber Science and Engineering, Southeast University, Nanjing, China.
Plos One
|June 12, 2026
Summary
A new dynamic defense framework enhances telemedicine security by integrating reinforcement learning, advanced data protection, and hardware acceleration. This system significantly boosts protection against cyber threats and improves real-time performance in remote medical settings.
Area of Science:
- Cybersecurity
- Medical Informatics
- Computer Science
Background:
- Telemedicine systems face increasing cybersecurity risks due to interconnectedness and widespread deployment of medical big models.
- Traditional static defenses are insufficient against rapidly evolving cyber threats in healthcare, leading to risks like model parameter leakage and privacy breaches.
- A dynamic defense framework is needed to balance security, privacy, and real-time performance in medical environments.
Purpose of the Study:
- To develop and evaluate a dynamic defense framework for enhancing the security and operational resilience of remote medical systems.
- To address cybersecurity challenges in telemedicine, including zero-day attacks, privacy leakage, and performance limitations.
- To test the hypothesis that intelligent decision-making, trusted coordination, and hardware acceleration can improve medical cybersecurity.
Main Methods:
- Developed a reinforcement learning (RL)-driven adaptive dynamic defense strategy as the core decision-making module.
- Integrated a security-enhanced model protection architecture using Shamir's threshold scheme, adversarial training, and differential privacy.
- Incorporated a blockchain-based verification mechanism with improved PBFT and FPGA-based hardware acceleration on the Xilinx XC7K325T platform.
- Evaluated the framework using NS-3, Python, PyTorch, Hyperledger Fabric, and the Synthetic IoMT Security Dataset.
Main Results:
- Increased the zero-day attack blocking rate from 68.5% to 99.3% in regional medical alliance and emergency ambulance scenarios.
- Improved medical image encryption throughput from 120 Mbps to 450 Mbps.
- Reduced CPU peak utilization by 47.8% and eliminated privacy leakage during cross-institutional data sharing.
- Stabilized core clinical service latency within 35ms.
Conclusions:
- The proposed dynamic defense framework significantly enhances the security and operational resilience of remote medical systems.
- The integration of RL, advanced data protection, blockchain, and hardware acceleration effectively addresses critical cybersecurity risks in telemedicine.
- The framework demonstrates a superior balance of security, privacy, and real-time performance for modern medical infrastructures.