Explainable zero-day attack detection in IoMT using transformer-based time-series modeling

Mahdee Jodayree, Arman Kavoosi Ghafi, Sara Amiri, Puya Shaykholeslami · 2026

Read Explainable zero-day attack detection in IoMT using transformer-based time-series modeling by Mahdee Jodayree, Arman Kavoosi Ghafi, Sara Amiri, Puya Shaykholeslami (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/pmc-13402581.txt.

Description

Mahdee Jodayree and colleagues designed an explainable transformer system to detect zero-day attacks on medical Internet of Things networks. Their 2026 paper tackles intrusion detection under severe class imbalance, unseen attacks, and deployment limits, where older methods rely on heavy feature work and fail on new datasets. A protocol-aware model could generalize better.

The team built a transformer with protocol-state masking at multiple scales that embeds protocol rules into attention, plus auxiliary heads and an energy-based open-set module. Trained on one IoMT benchmark and tested on a fully separate testbed, it reached about 99.6 percent accuracy on the unseen data and strong zero-day detection, with an AUROC of 0.975. The paper supports protocol-aware modeling for hospital networks, while live deployment testing remains ahead.

Catalog record

CREATORMahdee Jodayree, Arman Kavoosi Ghafi, Sara Amiri, Puya Shaykholeslami
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE107 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13402581

Browse the full library · About the Tower of Babel