autoSCORE

by Holberg EEG  · Based in Norway → — Developing intelligent software from uniquely structured brain data
Neurology

Overview

autoSCORE, developed by Holberg EEG AS, is a software-only decision support product featuring an AI algorithm trained on over 30,000 expertly annotated EEGs. It performs with expert-level accuracy in differentiating between normal and abnormal EEG recordings and sub-categorizing various abnormalities. The tool assesses the probability of abnormalities in previously acquired EEG sections and classifies them into specific types, including focal epileptiform, generalized epileptiform, focal non-epileptiform, and diffuse non-epileptiform. Designed to assist qualified medical practitioners, autoSCORE integrates seamlessly with compatible EEG review software, such as Natus NeuroWorks EEG software, to indicate potential abnormality markers. It aims to accelerate EEG review, reduce reporting variability, improve assessment quality, and enhance overall productivity by providing AI-driven insights without a direct user interface, allowing neuro care teams to focus more on patient care.

Reviewed by Pouyan Golshani, MD — Interventional Radiologist

Key Features

  • AI algorithm trained on over 30,000 annotated EEGs
  • Expert-level EEG interpretation, comparable to human experts
  • Flags clinically relevant abnormalities in EEG recordings
  • Provides study-level insights across LTM, ambulatory, and routine studies
  • Automates routine assessments and highlights key findings
  • Classifies abnormalities into four subtypes: focal epileptiform, generalized epileptiform, focal non-epileptiform, diffuse non-epileptiform
  • Provides visual preview of abnormality prevalence before opening a study
  • Helps navigate to abnormal EEG segments based on type and probability score
  • Seamlessly integrated into Natus NeuroWorks EEG software
  • Acts as a decision support tool to aid clinical workflow

Use Cases

  • Streamlining EEG review and interpretation
  • Reducing EEG reporting variability and improving quality
  • Reducing time and cost for EEG users
  • Assisting neurologists in the assessment of EEG
  • Supporting timely clinical decisions in neuro care
  • Enhancing productivity in high-volume EEG environments

What Physicians Need to Know

Evidence Base
autoSCORE is built upon the SCORE database, which is described as the world's largest repository of structured, highly annotated EEGs, collected over 10 years from hospitals using SCORE software. This database contains approximately 74,914 EEG records, with 43,845 being structured and labeled, including metadata such as medical diagnoses and medications. The SCORE-AI model, which powers autoSCORE, was developed using this database and trained on over 30,000 expertly labeled EEG recordings. The development and validation of the model were designed to avoid bias and focus on key aspects of EEG interpretation.
Clinical Validation Studies
autoSCORE has undergone clinical validation, demonstrating performance comparable to human experts in interpreting EEGs. A multicenter diagnostic accuracy study developed and validated the SCORE-AI model using EEGs recorded between 2014 and 2020. The model was trained on 30,493 EEG recordings annotated by 17 experts and validated using three independent test datasets comprising 9,945 EEGs. A 2023 study published in JAMA Neurology reported that autoSCORE consistently achieved accuracy, sensitivity, and specificity near or above 90%, performing on par with leading human experts and outperforming similar AI models. The algorithm is designed to distinguish normal from abnormal EEGs and further classify abnormal recordings into four clinically relevant categories: focal epileptiform, generalized epileptiform, focal non-epileptiform, and diffuse non-epileptiform.
Differential Diagnosis Support
autoSCORE supports differential diagnosis by classifying abnormal EEG recordings into four clinically relevant categories: focal epileptiform, generalized epileptiform, focal non-epileptiform, and diffuse non-epileptiform. This categorization provides important information for medical decisions regarding patient management.
Clinical Workflow Integration
autoSCORE is designed for seamless integration into existing clinical workflows. It is exclusively compatible with Natus NeuroWorks EEG software (version 10 GMA5 or higher) and BRAIN QUICK Software (version 4.02 or higher), which are widely used EEG platforms. The integration allows for an intuitive user experience, aiming to improve efficiency and time to diagnosis. It provides study-level visual indicators to support workday planning and helps quickly identify clinically significant traces or segments.
Decision Audit Trail
The provided information does not explicitly detail a decision audit trail feature for autoSCORE. However, general principles of AI audit trails suggest they should record inputs, outputs, model versions, confidence levels, and any human involvement to ensure accountability and compliance.
Physician Tip

Physicians should use autoSCORE as an AI-driven second opinion tool to assist in EEG interpretation, recognizing that it provides expert-level analysis and flags clinically relevant abnormalities. It can help prioritize critical cases and streamline the review of high volumes of EEGs, potentially reducing workload and improving efficiency. While autoSCORE offers comprehensive EEG interpretation, the physician's clinical judgment remains essential for making final diagnostic conclusions.

autoSCORE is exclusively integrated with Natus NeuroWorks EEG software (version 10 GMA5 or higher) and BRAIN QUICK Software (version 4.02 or higher). This integration allows autoSCORE to run on the local EEG system, ensuring that it does not transmit clinical data over networks or collect protected health information.

Details

Category Clinical Decision Support & Reference, Neurology AI
Pricing Unknown
Compliance
BAA AvailableUnknown
HIPAA CompliantUnknown
FDA Status Unknown
Integrations
EHR Not specified
Specialties Neurology

FDA clearance history

Each FDA authorization for this product and its earlier versions, newest first. From FDA’s list of AI-enabled medical devices; each number links to FDA’s record.

Decision dateNumberName as listed by FDA
2025-04-09K243743 510(k)autoSCORE
2024-01-07K231068 510(k)autoSCORE

Ratings & Reviews

No reviews yet. Be the first to review this tool!

Rate autoSCORE

Clinical Value
Ease of Use
Integration
Support & Docs
Value for Money
No press coverage or publications on file yet. Know of some? Suggest an edit →

Videos

Product demos, reviews, and walkthroughs for autoSCORE.

View all on YouTube

Frequently Asked Questions

autoSCORE is an AI-based model that provides expert-level review of EEG data, guiding physicians to studies and parts of studies requiring more in-depth assessment. It differentiates normal from abnormal EEG reads in both adult and pediatric studies and sub-classifies abnormal EEGs into four clinically relevant categories: Epileptiform – focal, Epileptiform – generalized, Non-epileptiform – focal, and Non-epileptiform – diffuse.
A current limitation of autoSCORE is that it was developed and validated on routine EEGs, excluding neonates and critically ill patients. Additionally, the model was trained to identify biomarkers visually recognized by human experts, rather than directly predicting diagnosis or therapeutic response.
Yes, autoSCORE has received FDA 510(k) clearance for routine EEG, long-term monitoring (LTM), and ambulatory EEG studies in the United States. It has also received the CE mark for European and other applicable countries.
While autoSCORE is described as the world's first AI model capable of automatic and comprehensive clinical EEG interpretation with accuracy on par with medical experts, a 2023 study in JAMA found that autoSCORE consistently outperformed other similar AI models on the market.
autoSCORE is exclusively available with Natus NeuroWorks EEG software, one of the most widely used EEG platforms. It seamlessly integrates into the software, automatically classifying and providing a visual summary of EEG abnormalities, flagging segments that may contain abnormalities to help guide prioritization and balance workloads.
autoSCORE was developed using a deep-learning model trained on the world's largest dataset of over 30,000 expertly labeled EEG recordings. This extensive training allows it to consistently identify abnormalities with accuracy comparable to highly trained clinicians and perform on par with leading human experts.
Yes, autoSCORE is designed to alleviate demands on clinical resources and staff by shortening the time to diagnosis and reducing physician workloads. It streamlines the review process by automating routine assessments and highlighting key findings, allowing neuro care teams to focus more on patient care.

Related Tools

Stanford University
Stanford University
Clinical Decision Support & Reference
Stanford Online's "AI in Healthcare" program provides a comprehensive overview of artificial intelligence's applications and implications within the medical field.
RADIN Health
RADIN Health
Clinical Decision Support & Reference
RADINA is an AI-powered platform by RADIN Health designed to optimize radiology workflows, enhance diagnostic accuracy, and streamline reporting.
Med-PaLM 2
Google
Clinical Decision Support & Reference
Med-PaLM 2 is a large language model from Google, fine-tuned for healthcare to provide high-quality answers to medical questions and support clinical decision-making.
Pacmed Critical Discharge
Pacmed
Clinical Decision Support & Reference
Pacmed Critical Discharge is an AI tool that predicts the probability of death or readmission within seven days of ICU discharge, providing explainable insights for physicians.
AiZtech Labs
AiZtech Labs
Clinical Decision Support & Reference
AiZtech Labs provides enterprise-grade AI solutions for healthcare, focusing on operational efficiency, diagnostic accuracy, and patient care workflows for providers and health systems.
Stethophone Pro
Sparrow Acoustics
Cardiology AI
Stethophone Pro transforms smartphones into advanced cardiac screening tools, enabling clinicians to capture, interpret, and visualize heart sounds for early detection of cardiovascular diseases.

See all Clinical Decision Support & Reference tools →

Suggest an Edit → | Report an issue | Last Verified: 2026-11-08 | First Added: 2026-04-18
AI Tool Finder
AI-powered search. Results may not be comprehensive.