Privacy-preserving adaptive arrhythmia monitoring with wearable ECG, smartphone PPG, and federated on-device learning
Aneta Kartali, Marija Koshtrevska, Stevan Jokić, Nenad Gligorić, Octavian M. Machidon · 2026
Read Privacy-preserving adaptive arrhythmia monitoring with wearable ECG, smartphone PPG, and federated on-device learning by Aneta Kartali, Marija Koshtrevska, Stevan Jokić, Nenad Gligorić, Octavian M. Machidon (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13461321.txt.
Description
Kartali and colleagues presented in 2026 Cardio-FL, a smartphone system for spotting heart rhythm faults outside clinics while keeping heart data private. Wearable ECG sensors and phone cameras can track rhythm at home, but a usable system must handle messy real-world signals, guard sensitive data, adapt to each user, and limit false alarms. Their prototype streams ECG from a chest wearable to an Android app for on-device analysis.
Kartali's model is tiny, with only about five and a half thousand parameters, and sorts beats into normal, early upper-chamber, and early lower-chamber types. The phone's camera adds a pulse check after alerts. Learning adapts on the device itself using unlabelled beats, and updates are shared in a federated way, meaning data stay on phones while model gains are pooled. Activity sensing gates when recording and training run, to save power and cut noise.
Catalog record
| CREATOR | Aneta Kartali, Marija Koshtrevska, Stevan Jokić, Nenad Gligorić, Octavian M. Machidon |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 80 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |