Labyrinthine Wrinkle-Patterned Fiber Sensors Based on a 3D Stress Complementary Strategy for Machine Learning-Enabled Medical Monitoring and Action Recognition.

Fiber strain sensors show good application potential in the field of wearable smart fabrics and equipment because of their characteristics of easy deformation and weaving. However, the integration of fiber strain sensors with sensitive response, good stretchability, and effective practical application remains a challenge. Herein, this paper proposes a new strategy based on 3D stress complementation through pre-stretching and swelling processes, and the polydimethylsiloxane (PDMS)/silver nanoparticle (AgNPs)/MXene/carbon nanotubes (CNTs) fiber sensor with the bilayer labyrinthian wrinkles conductive network on the PU fiber surface is fabricated. Benefiting from the wrinkled structure and the synergies of sensitive composite materials, the fiber sensor exhibits good stretchability (>150%), high sensitivity (maximum gauge factor is 57896), ultra-low detection limit (0.1%), fast response/recovery time (177/188 ms) and good long-term durability. It can be used as Morse code issuance and recognition to express the patient's symptoms and feelings. Further, the sensor enables comprehensive human movement monitoring and collects data of different characteristics with the assistance of machine learning, different letters/numbers are recognized and predicted with an accuracy of 99.17% and 99.33%. Therefore, this fiber sensor shows potential as a new generation of flexible strain sensors with applications in medical monitoring and human-computer interaction.