EEG-Based AI Functional Readouts Anchoring Translational NAM Platforms

Risorius Co., Ltd. develops deep learning–based EEG analytics that transform non-invasive human brain signals into interpretable, clinically relevant functional biomarkers. Our current work includes more than 10 retrospective clinical EEG and PSG studies conducted with hospitals in Korea and other Asian countries, focusing on feasibility, feature discovery, and optimisation of explainable AI pipelines.

In the context of HORIZON-HLTH-2026-01-TOOL-03, Risorius provides a human-functional brain readout layer that anchors NAM platforms to clinically observable outcomes. While our current technology is centred on human EEG/PSG data, it is explicitly designed to interface with other NAM components developed within the consortium.

Within the project, we will:

• Develop AI-driven functional brain biomarkers from human EEG/PSG data that capture intervention- and disease-related brain-state changes

• Enable alignment between human functional outcomes and preclinical or in vitro readouts generated by other consortium partners, supporting translational interpretation

• Define a scientifically grounded roadmap for cross-domain and cross-species translation, where relevant data are available within the consortium, without presupposing standalone animal-model replacement

By positioning human brain function as a clinically grounded reference point, Risorius supports the integration, interpretability, and translational value of NAM ecosystems, while maintaining a realistic and credible development pathway.