Diagnosis (general) · Diagnosis and imaging
Epilepsy patients texted their brain monitor, and false alarms dropped
In a Nature Medicine pilot, 13 patients in an epilepsy unit could chat with an AI linked to their live brain recordings and mark alarms as real or false. In a simulation using feedback from the patients who annotated alerts, the median false-alarm rate fell from 0.74 to 0.10 an hour. Rating: Early research.
How far away is this?
A small pilot with 13 people, and the false-alarm result is from a simulation; the real detector wasn't changed during stays.
Get the weekly How Far
One email every Friday morning (Sydney time) on which medical AI actually reached patients, and what it means for you. Free, in plain English.
Where is it approved?
| Region | Regulator | Status | Notes |
|---|---|---|---|
| United States | FDA | Not applicable | Research prototype. |
| European Union | CE mark / EMA | Not applicable | No register record checked. |
| United Kingdom | MHRA | Not applicable | No register record checked. |
| Australia | TGA | Not applicable | No ARTG record checked. |
Research prototype only. "Register not searched yet" means we haven't searched that regulator's register; it doesn't mean the product is or isn't available. Approval in one country isn't approval in another.
What is it?
Hospital EEG monitors watch for seizures but cry wolf a lot. This phone app let patients ask about their own brain activity and tell the system when an alert was wrong.
What's the evidence?
The study involved 13 patients in an epilepsy monitoring unit at Penn, using a smartphone AI chat connected to live EEG.[1]
Evidence stage: Early research · #c1
In a simulation afterwards, patient feedback cut median false alarms from 0.74 per hour to 0.10 per hour. Put simply: from roughly one false alarm every 80 minutes to one every 10 hours.[1]
Evidence stage: Early research · #c2
What does it mean for me?
If you or a family member has a hospital EEG stay, you can ask how alarms are checked. This app isn't in routine use.
What could I ask my doctor?
- How often do false alarms happen on this monitor?
These are general starting points, not a recommendation to choose any particular test or treatment.
Short videos, under a minute
We make short, plain-English videos about AI in medicine, including what it can't do yet. Follow How Far for the next one.
Change history
- : First published.
- : Editor pass before publication: labels aligned to site categories; sources and rating reviewed.
General information only, not medical advice. Talk to your doctor about your situation. AI-assisted source check. Not clinically reviewed.
References
- Nature Medicine (2026): Human–AI collaboration for comanaging epilepsy in real time. doi.org (2026). Peer-reviewed study. https://doi.org/10.1038/s41591-026-04680-4
- NIH news release. nih.gov. Press release (claim). https://www.nih.gov/news-events/news-releases/patients-engage-epilepsy-monitor-real-time-improving-seizure-detection
Publisher shown as the website hosting each source. Dates are given where the source label includes them.