Page 42 - Read Online
P. 42

Ma et al. Soft Sci 2024;4:26  https://dx.doi.org/10.20517/ss.2024.20             Page 19 of 34

               Table 3. Summary of machine learning-assisted soft biophysical sensors based on LIG for intelligent healthcare
                LIG composites         Signals Performance   Intelligent applications                 Ref.
                (role/substrate
                LIG (E)/PDMS           Pressure NA           Distinguish ten braille numbers with a high accuracy of 96.12%  [101]
                                                          -1
                LIG (E)/PU             Pressure Sensitivity: 0.05 kPa ;   Classify various voices with a high accuracy of 94.6%  [102]
                                             stability: 1,000 cycles
                LIG (S)/PDMS           Strain  GF: 2,336;    Clinical evaluation of actual CVD events, high accuracy of   [3]
                                             limit: 0.0056%;   98.7%
                                             clinical accuracy: >93%

               LIG: Laser-induced-graphene; PDMS: polydimethylsiloxane; PU: polyurethane; GF: gauge factor; CVD: cardiovascular disease.


               Meanwhile, Zhu et al. reported a soft non-enzymatic glucose sensor based on LIG-electrodes coated with a
               uniform metal (Ni or Ni/Au), which was packaged by a porous encapsulating reaction cavity for wearable
               applications [Figure 10E] . As expected, the fabricated glucose sensor based on LIG foams exhibited a high
                                    [40]
                                      -1
                                          -2
               sensitivity of 1,080 μA·mM ·cm , whereas the device with LIG fibers enhanced the detection sensitivity up
               to 3,500 μA·mM ·cm . Then, they attached the developed glucose sensor onto the arm surface to evaluate its
                             -1
                                 -2
               on-body performance. The results illustrated that the proposed glucose sensor could monitor the variations
               in glucose concentration in sweat, i.e., decreased from 0.26 to 0.15 mM, responding to 1 and 3 h after lunch
               [Figure 10F].
               Zhao et al. reported an iron nano-catalysts (FeNCs)/LIG-based wearable electrochemical patch, where
                                                                                                     [103]
               FeNCs functioned as the catalyst and LIG was utilized as the signal-improving substrate [Figure 10G] . A
               two-channel hydrogel chip was integrated with FeNCs/LIG to implement the wearable configuration.
               Profited from the unique 3D microstructures of LIG and the remarkable electrocatalytic activity of FeNCs,
               the wearable electrochemical patch exhibited outstanding sensing performance for sweat metabolites
               [tyrosine (Tyr) and uric acid (UA)] detection. When mounted onto a human skin surface, the device could
               accurately monitor Tyr and UA in sweat, essential in non-invasive health monitoring and management. The
               results exhibited that the concentrations of Tyr evaluated by the skin-integrated patch were comparable to
               those assessed by the liquid chromatography-mass spectrometry (LC-MS), illustrating good reliability
               [Figure 10H].


               Torrente-Rodríguez et al. reported a standalone wireless health management device, termed a graphene-
               based sweat stress monitoring system (GS4), to explore the dynamics of the sweat stress hormone
               [Figure 10I] . Combining LIG and immunosensing approaches effectively detected cortisol in human
                         [43]
               sweat and saliva. In addition, the GS4 device was attached to the skin surface to analyze sweat cortisol
               during a 50-minute stationary cycling exercise. The results revealed that the sweat cortisol increased and
               reached the highest level after continuous biking [Figure 10J]. This soft and wearable point-of-care device
               provided a practical approach for stress monitoring and continuous evaluation of the psychological states of
               subjects.


               In addition to biofluids, the exhaled gas carries essential information for disease diagnostics. For instance,
               exhaled NOx is a crucial biomarker for human respiratory disease diagnostics [104-106] . Motivated by this, Yang
               et al. developed a moisture-resistant and stretchable NOx gas sensor based on porous LIG, where the LIG
               sensing/electrode region was sandwiched between a semipermeable PDMS film and a soft elastomeric
                                                         [39]
               substrate for patient breath analysis [Figure 10K] . The gas sensor fabricated with optimized processing
               parameters showcased excellent performance: significant sensitivity of 4.18‰ ppm  to NO and 6.66‰
                                                                                        -1
                   -1
               ppm  to NO , and ultralow detection limit of 8.3 ppb to NO and 4.0 ppb to NO . The introduction of a LIG
                          2
                                                                                  2
               electrode with a serpentine pattern and strain isolation from the PI island enabled the gas sensor to undergo
   37   38   39   40   41   42   43   44   45   46   47