Fraunhofer-Institute for Integrated Circuits IIS and Sonovum GmbH

Sonovum GmbH developed a non-invasive intracranial pressure monitoring device based on a novel proprietary biosignal and AI-based classification models. Through TEF-Health, Fraunhofer IIS provided access to its Multimodal Biosignal Laboratory for controlled physical testing and feasibility studies. The collaboration refined software and machine learning components, established performance evidence, clarified regulatory pathways, and advanced the solution from TRL 6 to TRL 8. Sonovum is now conducting a clinical study to progress toward TRL 10.

SME / Company Sonovum GmbH

Service Provider Fraunhofer Gesellschaft Zur Forderung Der Angewandten Forschung Ev (Fraunhofer)

Disease / Clinical Domain Neurology

Product / Solution Non-invasive intracranial pressure monitoring device based on a novel proprietary biosignal and AI-based classifier models to interpret brain pulse curve data.

Validation Needs Controlled physical testing in a clinical setting Feasibility studies to establish performance evidence Methodological consulting to refine software and machine learning algorithms Interpretation of concurrent regulatory frameworks Validation of clinical data for a novel biosignal without predicate device

About Sonovum GmbH

Sonovum GmbH is a German medtech start-up developing a non-invasive device to measure intracranial pressure in patients suffering from severe brain trauma injuries. The system is based on CE-marked medical devices (TRL 10) that monitor the brain pulse curve and integrates AI-based classification models to interpret collected biosignal data.

The Challenge

Sonovum faced significant challenges due to the novelty of its proprietary biosignal and the absence of an existing predicate medical device. The path to validate clinical data was sophisticated by the complexity of the innovative approach and the necessity to interpret concurrent regulatory frameworks.

Additionally, Sonovum aimed to strengthen expert-based curation and annotation with a data-driven approach. The company required controlled clinical testing, performance validation, and methodological support to refine its software and AI classifier models while clarifying regulatory pathways.

The Story

Sonovum GmbH, a German medtech start-up, set out to develop a non-invasive device to measure intracranial pressure in patients suffering from severe brain trauma injuries. Based on the CE marked medical devices (TRL 10) which monitors the brain pulse curve. Despite its potential, the system faced significant challenges: The technology relied on a novel, proprietary biosignal with no existing predicate medical device. Additionally, the path to validate the clinical data was sophisticated by the complexity of the innovative approach, the necessity to interpret concurrent regulatory frameworks and because Sonovum wanted to underpin expert-based curation and annotation with a data-driven approach.

Article content
Through TEF-Health, Fraunhofer IIS offered access to its “Multimodal Biosignal Laboratory” service, where experts conducted controlled physical testing and feasibility studies in a clinical setting. The collaboration combined scientific communication support with methodological consulting to refine the software of the device and its machine learning algorithms. These tests established performance evidence, clarified regulatory pathways, and strengthened the technology’s credibility.

As a result, Sonovum’s AI-based classifier solution progressed from Technology Readiness Level (TRL) 6 to TRL 8, demonstrating measurable advancement toward real-world application. The device now integrates AI models to interpret the collected data and continues to evolve through iterative testing. Sonovum is currently conducting a clinical study that enables the final step from TRL 9 to TRL 10.

This partnership highlights TEF-Health’s mission in action: enabling European SMEs to bridge scientific ambition with clinical reality. By grounding innovation in evidence and expert collaboration, physical testing becomes the foundation of trust in healthcare AI.