Ceribell Clarity

AI-Driven EEG Diagnostics for Rapid and Informed Neurologic Care

Ceribell with Clarity® AI brings trusted, real-time EEG diagnostics to the bedside, ensuring fast and accurate detection of status epilepticus, empowering clinicians to make decisive and effective treatment choices

Clarity, the Ceribell machine learning algorithm, continuously interprets a patient’s EEG waveforms and displays actionable insights regarding seizure activity on the recorder, including automatic alerts in the event of non-convulsive status epilepticus.

First and only FDA-cleared AI technology for detection of electrographic seizures in all ages of patients, from preterm neonates through adults.1,2
Continuous EEG monitoring and seizure burden display

Remotely, Clarity provides:

Seizure burden display with pre-annotations to support efficient EEG interpretation

How does Clarity Work?

Clarity is a machine-learning, cloud-based AI algorithm that interprets the EEG every 10 seconds across all EEG channels to provide a seizure burden estimate.3

Seizure burden is the amount of seizure activity detected in the prior 5-minute window. Prolonged seizures, such as status epilepticus, which is associated with poor neurological outcomes, display a high seizure burden.

Key Features and Capabilities

  • Automatically detects suspected status epilepticus at the bedside
  • Continuously monitors the EEG for seizure burden
  • Trained on thousands of EEG annotated by epileptologist experts and is continuously improving
  • Studies generally demonstrate that the Clarity algorithm is specific and sensitive in detecting non-convulsive status epilepticus3,4

Practical Considerations

  • EEG offers one component of diagnostic insight. The decision to administer treatment, or not, should consider the combination of clinical history, patient presentation, and objective data.
  • Inter-rater variability of EEG diagnoses is well documented among expert readers. Since Clarity AI is trained based on expert input, a degree of inter-rater variability between Clarity AI’s output and the interpretation of the same case by a skilled neurologist should be expected.
  • The capability and accuracy of the Clarity AI continues to improve with increased training on a growing volume of cases.

Studies show that Clarity is highly accurate3,4

Clarity is validated to consistently alert for suspected status epilepticus and rule-out seizure4 – on par with top epileptologists7

Learn More
decorative

References and Citations

1. Ceribell, Inc., (2025, April 15) [Press release]
2. Ceribell, Inc., (2025, November 24) [Press release]
3. Kamousi, B., et al. (2021) Neurocrit Care. 34(3):908-917
4. Puranik, T., et al., (2025). Neurocritical Care Society Annual Meeting Poster
5. De Marchis, G.M., et al. (2016) Neurology. 86(3):253-260
6. Payne, E.T., et al. (2014). . Brain. 137(Pt 5):1429-1438
7. Internal analysis on the Puranik, T., et al., (2025). Neurocritical Care Society Annual Meeting Poster dataset. Individual reviewer’s performance was estimated on each subset reviewed. The final Clarity performance reported is the average across all the subsets