HCAP: Hybrid cyber attack prediction model for securing healthcare applications

Area of Science:

• Cybersecurity
• Healthcare Technology
• Machine Learning

Background:

• The Internet of Medical Things (IoMT) enhances healthcare but introduces cybersecurity vulnerabilities.
• Attacks threaten sensitive health data confidentiality, integrity, and availability.
• Existing models face computational complexity and overfitting issues.

Purpose of the Study:

• To enhance cyberattack prediction and prevention in IoMT environments.
• To improve healthcare data security and IoMT system resilience using machine learning.
• To address dataset availability and model efficiency limitations.

Main Methods:

• Developed a hybridized cyber attack prediction (HCAP) model.
• Utilized Principal Component-Recursive Feature Elimination (PC-RFE) for feature selection.
• Employed a lion-optimization technique with Long Short-Term Memory (LSTM) networks for prediction.

Main Results:

• The HCAP model achieved 98% accuracy in detecting cyberattacks.
• Reduced false positive rates by 25% and false negative rates by 20%.
• Demonstrated a 30% improvement in computational efficiency compared to existing models.

Conclusions:

• The HCAP model offers a reliable and efficient solution for IoMT cybersecurity.
• Enhanced threat detection capabilities improve the overall security of healthcare applications.
• The study highlights the potential of advanced machine learning in securing critical healthcare infrastructure.