Ceribell Clarity AI links seizure burden to neurological outcomes
Ceribell has announced the publication of a new peer-reviewed study in Critical Care Medicine showing that its Clarity artificial intelligence algorithm's measure of seizure burden is correlated with patients' functional outcomes at hospital discharge. The findings, drawn from a multisite dataset of 359 adult patients, add prospective clinical weight to the company's point-of-care electroencephalography (POC EEG) platform.
The study was conducted across three academic centres — Yale School of Medicine, Massachusetts General Hospital, and the University of New Mexico School of Medicine — where both the Ceribell POC EEG system and near-continuous conventional EEG were simultaneously available. Researchers analysed the relationship between seizure burden as quantified by the Clarity algorithm and modified Rankin Scale (mRS) scores at discharge, a widely accepted functional outcome measure.
What the data show
The headline finding is a 3.4-fold increase in the odds of death or severe disability (mRS ≥ 4 at discharge) for patients whose peak five-minute seizure burden reached 90% or above at any point during monitoring, compared with those who recorded no seizure burden. Equally striking is a near-doubling of risk for each additional hour of Clarity-detected seizure activity, with an adjusted odds ratio of 1.98. The authors describe this as a "dose-response" relationship — a meaningful statistical pattern that strengthens causal interpretation, though the study's retrospective design means definitive causality has not been established.
Josef Parvizi, co-founder, board member and Chief Medical Adviser at Ceribell, and lead author of the paper, said the research "confirms that Ceribell's Clarity AI algorithm is detecting biomarkers of a disease state that are clinically relevant, providing a window into the patient's prognosis."
The study builds on the SAFER-EEG retrospective multicentre study published in Neurology in July 2024, which first characterised the algorithm's comparative effectiveness against conventional EEG. The new Critical Care Medicine paper extends that work to patient-level functional outcomes at discharge.
Market and competitive context
Point-of-care neurological monitoring is an emerging segment within broader digital health and medical-device markets. Conventional EEG interpretation requires specialist neurologists and is operationally ill-suited to the emergency department or ICU environment, where patient deterioration can be rapid and specialist availability is inconsistent. Ceribell's FDA-cleared platform — cleared for seizure and delirium detection in ICUs and emergency rooms — addresses this gap by combining portable hardware with AI-driven interpretation at the bedside.
A growing number of academic hospital systems and commercial entrants are exploring AI-assisted EEG reading, and larger monitoring and neurophysiology players have begun acquiring capabilities in this space. Ceribell's competitive positioning rests on the depth of its clinical validation dataset and the specificity of its FDA clearance for acute-care settings — the new Critical Care Medicine publication strengthens both pillars.
For Ceribell, which listed on the Nasdaq as CBLL, the publication also carries commercial significance. Demonstrating a quantified outcome benefit — not merely technical equivalence to conventional EEG — is the kind of evidence that supports reimbursement conversations with payers and adoption decisions by hospital procurement committees. The company has not yet disclosed any payer coverage milestones or health-technology assessment submissions, and clinicians will look for prospective randomised data before drawing firm conclusions about whether reducing Clarity-measured seizure burden translates directly into improved outcomes at scale.