IntruSafe: Guarding Hospital Devices Against Text and Image Attacks

By Nuruzzaman Faruqui

·

Connected medical devices are cheap, low-powered and full of valuable health data, which makes them a magnet for cybercriminals. IntruSafe is an intrusion detection system that protects medical networks from both malicious text and malicious image data.

97.66%detection accuracy
98.50%precision
0.18%service interruption rate

The problem

The Internet of Medical Things is booming, but manufacturers keep prices low by using simple architectures that leave little room for sophisticated security. Healthcare data is valuable and confidential, so intrusions are growing fast.

What the research did

IntruSafe combines a Fully Connected Convolutional Neural Network (FCNN) with Long Short-Term Memory (LSTM) networks in a “sandwich” architecture. Together they detect malicious signals in both text-like network data and 2D image data, and act to stop further intrusion.

What we found

IntruSafe detected intrusions with 97.66% accuracy, 98.50% precision, 97.33% recall and a 97.85% F1-score, while interrupting normal service only 0.18% of the time.

Who benefits

  • Hospitals and clinics
  • Patients using connected medical devices
  • Medical device manufacturers

Limits and what comes next

Evaluated on research datasets; real hospital networks and evolving attacks require ongoing testing and updates. It builds on our earlier SafetyMed system.

The team

This research was led by Moutaz Alazab and Albara W. Awajan (Al-Balqa Applied University) with Areej Obeidat (Deakin University), me and Hafeez Ur Rehman (FAST-NUCES, Pakistan).

Frequently asked questions

What makes IntruSafe different?

It detects attacks in both text-like network data and image data using a combined FCNN–LSTM design.

Does it disrupt normal device use?

Only minimally: a 0.18% service interruption rate in testing.

How does it relate to SafetyMed?

Both protect medical IoT; IntruSafe extends the approach with a new sandwich architecture.

About the author

Nuruzzaman Faruqui is an award-winning AI educator and researcher who has taught artificial intelligence and machine learning at university level since 2018. A two-time winner of the Excellence in Teaching Award (2023, 2024) and winner of the Best Researcher Award (2024) at Daffodil International University, he is now a doctoral researcher in AI at Monash University, Australia. He has published 53 papers and six books, and writes these articles to make AI research easy for everyone to understand. His teaching · About Zaman · Get in touch

Keep reading