Artificial Intelligence–Enhanced Rapid Response Electroencephalography for the Identification of Nonconvulsive Seizure
Bibliographic record
Abstract
What Is the Issue? A nonconvulsive seizure is seizure activity defined by an altered mental status, subtle limb twitches, or changes in speech. They are more difficult to identify than convulsive seizures as they do not have the distinctive motor activity associated with convulsive seizures. Patients in emergency departments (EDs) and intensive care units (ICUs) with suspected nonconvulsive seizures must be monitored with an electroencephalogram (EEG) to confirm diagnosis. Rapid detection of nonconvulsive seizures is crucial — a delay in treatment risks brain injury. There can be significant delays in accessing conventional EEG monitors and treatment because of limited supply in critical care settings, such as EDs and ICUs. What Is the Technology? The Ceribell system is a rapid response point-of-care EEG designed for use in the ED and ICU to help identify patients who are having nonconvulsive seizures. The portable device has an artificial intelligence (AI) algorithm, Claritγ, that monitors seizure activity within a 5-minute interval to determine the seizure burden during that time frame. The device alerts a bedside care provider if seizure activity occurs. This information can guide physicians’ treatment plans. What Is the Potential Impact? Within the ED and ICU, Ceribell could be used to increase access to EEG, allowing for faster detection of nonconvulsive seizures. Conventional EEG monitors are expensive and usually require a trained specialist to use and interpret findings. Most hospitals have limited access to conventional EEGs, which can lead to delays in treating patients with nonconvulsive seizures. In critical care settings, Ceribell may improve efficiency and patient flow by shortening time to diagnosis, preventing unnecessary treatment escalation, decreasing transfers to tertiary care hospitals. What Else Do We Need to Know? The Ceribell system is not available in Canada as of this writing. The Ceribell system could improve time to treatment for patients with suspected nonconvulsive seizures due to the complexity and personnel needs of conventional EEG systems. Current research on the Ceribell system has mostly been retrospective with small sample sizes; therefore, the results may not be generalizable to a wider population. If the system is implemented in hospitals, training is required to ensure that all health care professionals in the ED or ICU know when to order the portable EEG so that the use of Ceribell is prioritized for patients with suspected nonconvulsive seizures. Datasets used to train AI algorithms tend to underrepresent equity-deserving groups, so implementing AI systems such as Ceribell with Clarity in health care settings could increase health care inequity for those not well represented in algorithm data.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".