Machine-Learning Mental-Fatigue-Measuring μm-Thick Elastic Epidermal Electronics (MMMEEE).

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Neuroscience

Background:

• Electrophysiological (EP) signals are vital for monitoring mental fatigue (MF) and overall health.
• Current wearable EP-based MF systems are often bulky and require personalized, labeled data.
• High-performance, ultrathin epidermal electrodes are needed for unobtrusive EP sensing.

Purpose of the Study:

• To develop a facile, scalable fabrication method for high-performance epidermal electrodes.
• To create plug-and-play wireless epidermal electronics for MF recognition.
• To implement an unsupervised transfer learning (UTL) scheme for accurate, user-independent MF monitoring.

Main Methods:

• A printing-welding-transferring (PWT) strategy was used to create micropatterned silver nanowire (AgNW) electrodes.
• AgNWs were welded via plasmonic effect, and electrodes were transferred to skin like tattoos.
• An unsupervised transfer learning (UTL) scheme was integrated with wireless multimodal epidermal electronics.

Main Results:

• The PWT strategy yielded conformable, comfortable, and stable electrodes for EP sensing.
• The developed system achieved accurate MF recognition across diverse users.
• The UTL scheme effectively minimized inter-subject differences without requiring user-specific labels or extensive computation.

Conclusions:

• A scalable PWT fabrication method enables high-performance, imperceptible epidermal electrodes.
• Integrated UTL with wireless epidermal electronics provides a user-friendly, accurate solution for MF monitoring.
• This approach overcomes limitations of current systems, paving the way for broader application of wearable EP sensing.