Secure pairing for implantable medical devices using vibration and motor-behavior entropy

• Perceptible vibration can be used for secure pairing between implantable medical devices (IMDs) and external devices, addressing security risks of imperceptible pairing.¹
• A proof-of-concept leverages natural randomness in human motor behavior as a shared source of entropy for key bootstrapping.¹
• The approach is potentially deployable to current IMD products (e.g. pacemakers, and by extension neuroprosthetics and BCI).¹
• The work is peer-reviewed in JMIR Biomedical Engineering and addresses regulatory and safety concerns as wireless implants scale.^{1 1}

Weekly enrichment (2026-07-20)

• The study (JMIR Biomedical Engineering, volume 10, article e57091) evaluated the prototype in a lab study of 24 participants (11 male, 13 female, ages 18-52), preceded by a 6-person pilot and a co-design workshop with a 74-year-old cardiologist and a 79-year-old pacemaker patient.²
• The pairing mechanism has the external device emit a fixed vibration (tested at 50/75/100 Hz for 400/700/1000 ms) while both the IMD and external device record z-axis accelerometer data; the FFT frequency-peak location serves as the shared entropy source rather than the vibration carrying the secret.²
• A single attachment motion carried between 2.61 and 4.48 bits of entropy, increasing with longer vibration duration and higher frequency.²
• The scheme uses a fuzzy extractor to reconcile the small mismatch between devices and then a password-authenticated key agreement (PAKE) to establish a 128-bit key; the mismatch threshold was set to guarantee a false acceptance rate (FAR) of 0.²
• The best-performing configuration (100 Hz, 700 ms, four attachment motions) achieved FAR = 0, false rejection rate (FRR) = 0.6%, roughly 15.5 bits of entropy, and completed in about 4.8 seconds.²
• An alternative condition of 50 Hz for 1000 ms with five attachments reached FAR = 0 and FRR = 3.7%.²
• Usability was high on the System Usability Scale, the extracted randomness passed the NIST statistical test suite (all p > .01), and 18 of 24 participants reported no discomfort.²
• Advantages over prior out-of-band pairing include lenient synchronization requirements and a low-sampling-rate accelerometer; the method targets deployable IMDs such as pacemakers and could extend to neuroprosthetics and BCI, though real IMD patients were not tested for ethics reasons.^{2 3}

Footnotes

1. https://biomedeng.jmir.org/2025/1/e57091 ↩ ↩² ↩³ ↩⁴ ↩⁵

2. https://doi.org/10.2196/57091 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. https://research.birmingham.ac.uk/en/publications/using-vibration-for-secure-pairing-with-implantable-medical-devic/ ↩