Open Access

ARTICLE

Deep Learning-Driven Anomaly Detection for IoMT-Based Smart Healthcare Systems

by Attiya Khan¹, Muhammad Rizwan², Ovidiu Bagdasar^2,3, Abdulatif Alabdulatif^4,*, Sulaiman Alamro⁴, Abdullah Alnajim⁵

1 Department of Computer Science, Kinnaird College for Women, Lahore, 54000, Pakistan
2 School of Computing, University of Derby, Derby, DE221GB, UK
3 Department of Mathematics, Faculty of Exact Sciences, “1 Decembrie 1918” University of Alba Iulia, Alba Iulia, 510009, Romania
4 Department of Computer Science, College of Computer, Qassim University, Buraydah, 52571, Saudi Arabia
5 Department of Information Technology, College of Computer, Qassim University, Buraydah, 52571, Saudi Arabia

* Corresponding Author: Abdulatif Alabdulatif. Email:

Computer Modeling in Engineering & Sciences 2024, 141(3), 2121-2141. https://doi.org/10.32604/cmes.2024.054380

Received 27 May 2024; Accepted 23 September 2024; Issue published 31 October 2024

Abstract

The Internet of Medical Things (IoMT) is an emerging technology that combines the Internet of Things (IoT) into the healthcare sector, which brings remarkable benefits to facilitate remote patient monitoring and reduce treatment costs. As IoMT devices become more scalable, Smart Healthcare Systems (SHS) have become increasingly vulnerable to cyberattacks. Intrusion Detection Systems (IDS) play a crucial role in maintaining network security. An IDS monitors systems or networks for suspicious activities or potential threats, safeguarding internal networks. This paper presents the development of an IDS based on deep learning techniques utilizing benchmark datasets. We propose a multilayer perceptron-based framework for intrusion detection within the smart healthcare domain. The primary objective of our work is to protect smart healthcare devices and networks from malicious attacks and security risks. We employ the NSL-KDD and UNSW-NB15 intrusion detection datasets to evaluate our proposed security framework. The proposed framework achieved an accuracy of 95.0674%, surpassing that of comparable deep learning models in smart healthcare while also reducing the false positive rate. Experimental results indicate the feasibility of using a multilayer perceptron, achieving superior performance against cybersecurity threats in the smart healthcare domain.

---

Keywords

---