SecureNeuroAI

SecureNeuroAI, funded by the German Federal Ministry of Research, Technology and Space (BMFTR) aims to develop legally compliant and secure AI-based solutions for real-time seizure detection from multimodal data - with a special focus on protecting against cyberattacks that could manipulate data or AI workflows and endanger patient safety.

Motivation

Medical devices and applications that use artificial intelligence (AI) have the potential to improve patients' quality of life. However, integrating AI into patient care increases the risk of cyberattacks, which can jeopardize patient safety and affect emergency services and medical device functionality.

Objectives and Approach

The SecureNeuroAI project aims to develop secure, AI-supported methods for real-time detection of medical emergencies, using epileptic seizures as an example. Detection is based on a comprehensive analysis and recording of multimodal sensor data. Portable electronic devices ("wearables") that record vital parameters, such as heart and respiratory rates, as well as clinical data, are used for this purpose. This data is analyzed by cyber-secure AI models designed to detect seizures and distinguish them reliably from potential data manipulation. Meanwhile, the project team is defining technical, organizational, and legal measures to support integrating these AI methods into clinical and domestic use cases.

Innovations and Vision

The project's results are an important step in strengthening the cybersecurity of critical medical devices that use AI methods to detect life-threatening conditions in real time. These new technical solutions protect both AI models and their underlying data from manipulation. The project results will be applied far beyond epilepsy seizure detection. Thus, the project creates a technological basis for significantly increasing the integrity, availability, and reliability of AI-based medical devices.