Pullulan-based semi-interpenetrating network hydrogel sensor for artificial intelligence-driven pressure recognition

Hydrogel sensors have gained significant attention in recent years due to their ability to detect various stimuli. This work first employs a "one-pot method" to construct a semi-interpenetrating network (sIPN) gel composed of polyvinyl alcohol (PVA) and pullulan polysaccharide (PUL), combined with potassium chloride (KCl) and MXene to establish an electrically conductive network. Then, the hydrogel is subjected to network contraction through solketal (SK) treatment to enhance its mechanical properties (named PKPM-S). After the network contraction, the mechanical properties of the PKPM-S hydrogel are improved approximately 9.6 times, with the maximum fracture strength of 5.706 MPa. Additionally, at a strain below 40%, the compressive performance of PKPM-S is enhanced by a factor of 2. The synergistic effect of MXene and KCl imparts high sensitivity to the PKPM-S hydrogel, with a gauge factor (GF) of 4.6 and a response time of 100 ms. Leveraging the superior sensing performance of PKPM-S and integrating wireless Bluetooth technology, we develop a wireless remote intelligent monitoring system for foot morphology and human movement. This system incorporates a convolutional neural network (CNN)-based deep learning algorithm for gait recognition and assistive rehabilitation functions. PKPM-S hydrogel shows significant potential for applications in medical rehabilitation and artificial intelligence (AI).