The rapid adoption of Edge Computing and Internet of Things (IoT) technologies in healthcare has enabled real-

time patient monitoring and low-latency clinical decision support. However, the distributed and resource-

constrained nature of Edge-IoT systems makes them highly vulnerable to cyber-attacks such as data breaches,

ransomware, and denial-of-service, which threaten patient privacy and system reliability. Traditional centralized

AI-based intrusion detection systems (IDS) face limitations in privacy preservation, scalability, and suitability

for edge environments. To address these challenges, this paper proposes a secure and privacy-preserving cyber-

attack detection framework that integrates an optimized LSTM Gated Multi-Layer Perceptron Neural Network

The integration of Edge Computing and the Internet of Things (IoT) has transformed modern healthcare by

enabling continuous patient monitoring, real-time data analytics, and rapid clinical decision-making. Wearable

sensors, smart medical devices, and edge nodes now collect and process sensitive health data close to the source,

reducing latency and improving responsiveness.

Despite these advantages, the decentralized architecture of Edge-IoT systems significantly increases their

exposure to cyber threats. Attacks such as unauthorized access, data manipulation, ransomware, and denial-of-

service can compromise the confidentiality, integrity, and availability of critical healthcare information, directly

affecting patient safety and trust.

Artificial Intelligence (AI) has emerged as a powerful tool for detecting and mitigating cyber-attacks by learning

Another challenge in intrusion detection is handling high-dimensional and noisy data. Effective feature selection

and optimization are essential to reduce redundancy, improve accuracy, and lower computational cost. Cyber-

attacks often exhibit temporal behaviour, requiring models that can learn both sequential and nonlinear patterns

Recent studies emphasize that cyber security in healthcare must address confidentiality, integrity, and

availability while operating under strict privacy regulations. Cybersecurity in healthcare IoT and edge computing

environments has become a critical research area due to the increasing number of cyber-attacks targeting

sensitive patient data. Researchers have explored AI, block chain, and federated learning to build secure, privacy-

preserving, and intelligent systems.

Goffer et al. propose an AI-enhanced cyber threat detection framework for critical infrastructure, emphasizing

deployment.Gabriel’s thesis focuses on ethical and privacy issues in AI-based healthcare data collection. It

concludes that decentralized data processing models such as federated learning can significantly reduce data

leakage risks [3].

Bashir et al. present a comprehensive study on Federated Learning in the Healthcare Metaverse, showing that

FL enables collaborative learning without exposing raw patient data. However, they note the need for secure

aggregation and audit mechanisms [4].

Singh and Sevukamoorthy integrate block chain with AI-based threat detection in financial networks. Their

results show improved integrity and tamper resistance, but scalability and latency remain open challenges [5]

.Mansourd evelops a block chain-enabled deep learning intrusion detection model using optimization and

recurrent networks. The study shows improved detection accuracy in cyber-physical systems, but it does not

for privacy-preserving collaboration [8]. Awotunde et al. use KPCA-CNN with block chain for smart city IoT

security. Their hybrid deep learning model improves prediction accuracy and data integrity, but feature

dependency and scalability are not fully optimized [9].

Researchers worked on various platforms using various Learning algorithms. The domains and techniques used,

key contributions, limitations are discussed as mentioned in the previous section.

Table1: Analysis of Literature Review

Deep Reinforcement Learning (DRL) models use reward mechanisms to interact with the environment and learn

the best course of action. By regularly revising its policy, DRL in intrusion detection adjusts to shifting attack

patterns. However, DRL's capacity to identify intricate temporal correlations in network traffic is constrained by

its usual reliance on centralized training and simple feature filtering. Table 1 illustrates that DRL is less

appropriate for real-time, privacy-sensitive healthcare IoT contexts due to its middling accuracy (77%) and

comparatively increased time complexity (0.79 ms).

An input layer, one or more hidden layers, and an output layer make up a feed-forward artificial neural network

known as a Multi-Layer Perceptron Neural Network (MLPNN).

1. Goffer, M. A., et al., “AI-Enhanced Cyber Threat Detection and Response: Advancing National Security

in Critical Infrastructure,” Journal of Posthumanism, vol. 5, no. 3, pp. 1667–1689, 2025.

2. Virk, A., et al., “Digital Health Policy and Cybersecurity Regulations Regarding Artificial Intelligence

(AI) Implementation in Healthcare,” Cureus, vol. 17, no. 3, 2025.

3. Gupta, S., Kamboj, S., and Bag, S., “Role of Risks in the Development of Responsible Artificial

Intelligence in the Digital Healthcare Domain,” Information Systems Frontiers, vol. 25, no. 6, pp. 2257–

2274, 2023.

4. Gabriel, O. T., Data Privacy and Ethical Issues in Collecting Health Care Data Using Artificial

Intelligence Among Health Workers, M.S. Thesis, Center for Bioethics and Research, 2023.

5. Bashir, A. K., et al., “Federated Learning for the Healthcare Metaverse: Concepts, Applications,

8. Karn, A. L., et al., “Applying the Defense Model to Strengthen Information Security with Artificial

Intelligence in Computer Networks of the Financial Services Sector,” Scientific Reports, vol. 15, no. 1,

9. Bhambri, P., and Starostka-Patyk, M., “Integrating AI with Blockchain for Decentralized Security and

Threat Prevention,” in Handbook of AI-Driven Threat Detection and Prevention, CRC Press, 2025, pp.

353–370.

10. Awotunde, J. B., et al., “Privacy and Security Enhancement of Smart Cities Using Hybrid Deep Learning-

Enabled Blockchain,” Scalable Computing: Practice and Experience, vol. 24, no. 3, pp. 561–584, 2023.

11. Konstantinos P. Ferentinos.,”Deep learning models for plant disease detection and diagnosis”

https://doi.org/10.1016/j.compag.2018.01.009

12. G.L. Grinblat et al. Deep learning for plant identification using vein morphological patterns Comput.

15. Too EC, Yujian L, Njuki S, Yingchun L. A comparative study of fine-tuning deep learning models for

plant disease identification. Computers and Electronics in Agriculture. 2019 Jun 1;161:272-279.

16. Moustafa, N., Slay, J“UNSW-NB15: A Comprehensive Data Set for Network Intrusion Detection

Systems”Military Communications and Information Systems Conference, 2015.

17. Shone, N., et al.“A Deep Learning Approach to Network Intrusion Detection.”IEEE Transactions on

Emerging Topics in Computational Intelligence, 2018.

18. Doshi, R., Apthorpe, N., Feamster, N.“Machine Learning DDoS Detection for Consumer Internet of

Things Devices.”IEEE Security and Privacy Workshops, 2018.

19. Kim, G., Lee, S., Kim, S.“A Novel Hybrid Intrusion Detection Method Integrating Anomaly Detection

with Misuse Detection.”Expert Systems with Applications, vol. 41, 2014.