top of page

Developing intelligent software from              uniquely structured brain data 

Holberg EEG has the world's most extensive collection of highly annotated EEG recordings. Our database offers a wealth of invaluable insights into brain activity patterns, used to automate clinical EEG analysis and empowering health insight for personal use

doctor and patient 1_edited.jpg


We are thrilled to announce the groundbreaking development of SCORE-AI, the world's first AI model capable of a fully automated and comprehensive assessment of routine clinical EEGs with human expert-level performance

Screenshot 2023-06-20 082750.png
Podcast live.png

Listen to the JAMA Neurology podcast where Dr. Sandor Beniczky talks about clinical opportunities with SCORE-AI

Editorial original.png

Read the JAMA Neurology Editors' choice

Holberg EEG offers standardization tools for EEG reporting and innovative AI-based products aiming to enhance reporting quality and efficiency. Our unique SCORE database is the foundation for our disruptive innovations, and we work globally to accelerate AI in healthcare through collaborations with the industry, academics, hospitals, and partners. We see great opportunities to expand our applications to personal health in wearables. 

                   Please get in touch to discuss how Holberg EEG can support your next project.



hiSCORE is our intuitive software application building on the harmonized standard, SCORE. It aims to provide quality enhancements in your EEG reporting, improve the overall clinical workflow, and strengthen your educational- and research opportunities.

SCORE-AI is an AI model developed using our database of highly annotated EEGs. SCORE-AI has the ability to distinguish abnormal from normal EEG recordings and classify abnormal EEGs into categories relevant for clinical decision making.  SCORE-AI archives human expert performance in fully automated interpretation of routine EEGs

The SCORE database has been collected from users of the hiSCORE software over a period of more than 10 years. The EEGs have been extensively labeled by highly qualified physicians and contain significant metadata and links to the EEG report and findings. The SCORE database has been used to develop the SCORE-AI model based on full-scale EEG with results achieving human expert performance.


We have now also developed a new AI model based on one-channel ear-to-ear EEG, showing remarkable performance promising for applications in the consumer market.


With our vision in mind, we will unite medical data and invite the industry, academics, hospitals, and partners to collaborate to develop new innovative tools to improve the value of care. 

Please get in touch to explore opportunities for data sharing and AI projects.   

bottom of page