Ultra-low bioelectronic impedance electrode based on molecule anchoring and preparation method

Abstract

The invention discloses an ultralow bioelectronic impedance electrode based on molecule anchoring and a preparation method, and relates to the field of bioelectronic impedance electrodes. In the production of the gel electrode, the generation and use of the ion conductive gel can securely anchor on the skin, thereby reducing the above barrier effects and increasing the ion-electron coupling area. Ion conductive gel is used in a gel electrode, and the electrode is used in an artificial limb device, so that when an amputee uses a myoelectric artificial limb, the signal quality is improved, the electrode and the skin are effectively coupled, the control precision of the artificial limb is optimized, physiological conflicts are reduced, and the comfort level of the artificial limb is improved; the amputee can control the artificial limb more accurately, and the use experience and life quality of the amputee can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of bioelectronic impedance electrodes, in particular to an ultra-low bioelectronic impedance electrode based on molecular anchoring and a preparation method thereof. BACKGROUND

[0002] Myoelectric control is an advanced technology for amputees that involves detecting, identifying, and applying undistorted myoelectric signals to control a prosthesis. However, due to long-term lack of activity leading to fibrosis and atrophy of the skin in amputees, the ionic-electronic coupling efficiency of the skin electrode interface is low, which easily distorts low-level myoelectric signals, especially signals with a maximum voluntary contraction (MVC) of less than 10%.

[0003] Dislike but dependence on powered prostheses is a typical feeling of amputees, and this dislike leads to an ultimate average rejection and abandonment rate of these prostheses of 35%. In order to realize the ultimate dream of amputees: man and prosthesis as one, people have been actively exploring to design a myoelectric prosthesis with comparable ability and comfort to human limbs. As a basic premise, undistorted high-fidelity myoelectric signals can ensure accurate identification of phantom limb activities, thereby achieving precise prosthesis control. Myoelectric signals are electromotive potentials generated when muscles contract. These signals contain neural information sent from the brain to control specific movements, and thus can be used as an input source for prosthesis control, generated by trained amputees. Myoelectric signals can be recorded by non-invasive electrodes placed on the skin of the residual limb. However, these signals vary greatly, usually standardized by the maximum voluntary contraction (MVC) of the muscle. For fine and delicate movements such as finger and wrist movements, muscle contraction is less than 10% MVC. The challenge is that the myoelectric signals generated by these movements are often distorted and have a very low signal-to-noise ratio (SNR) due to the inability of traditional electrodes to effectively couple with the skin of amputees. Therefore, amputees often have to use stronger muscle contractions (MVC exceeding 10%) to trigger the identification of weak but fine finger movement signals. This mismatch between the muscle contraction of the amputee and the movement of the prosthesis creates a physiological conflict between the new reflex arc and the limb muscle memory. It also creates severe incoordination and excessive friction at the interface of the prosthesis and the residual limb, leading to discomfort, excessive sweating, skin allergies, and ultimately abandonment of myoelectric prostheses. To solve this mismatch problem and allow amputees to control prostheses more accurately, improving the quality and availability of low-level myoelectric signals, especially signals below 10% MVC, is crucial to compensate for this mismatch.

[0004] The detection of myoelectric signals involves the capacitive coupling of ionic flux in the skin and electronic current in the gel electrode through two electrical double layers (EDL) on the skin-electrode interface. Traditionally, the methods to enhance the ionic-electronic coupling and reduce the bioelectronic impedance are mainly to make the electrode more flexible and stretchable, or to improve the interaction between the electrode and the skin surface molecules to adapt to the curved skin. However, these methods can only improve the detection of low-level signals in normal skin. The capacitive coupling of amputated skin is still weak because of the changes inside the skin, such as fibrosis and atrophy caused by amputation. This causes the imbalance of intracellular and extracellular calcium concentration in amputated skin due to long-term inactivity, thereby reducing the induced charge of the electrode metal layer in the capacitive coupling process. This imbalance raises the interface impedance level from a value close to the inherent impedance of the human body (6-10 kΩ) to from high kΩ to low MΩ, thereby failing to record myoelectric signals below 10% MVC with high quality (which is crucial to compensate for this mismatch). It is known that about 80% of the bioelectronic impedance (i.e. the so-called barrier effect) comes from the outermost layer of the skin - the stratum corneum (SC). SUMMARY

[0005] The present application aims to overcome one or more of the deficiencies of the prior art and provide an ultra-low bioelectronic impedance electrode based on molecular anchoring and a preparation method.

[0006] The purpose of the present application is achieved by the following technical solutions:

[0007] A method for preparing an electrode with ultra-low bioelectronic impedance, comprising:

[0008] S1. Making a PDMS substrate, mixing liquid PDMS with PMDS curing agent at a ratio of 10:1, and then stirring the mixture to uniformity using a centrifugal stirrer;

[0009] S2. After the mixture is uniform, place it in a vacuum chamber to eliminate air bubbles, and then deposit a PDMS layer on a fluorinated silicon wafer;

[0010] S3. The method for making a fluorinated silicon wafer is to immerse the silicon wafer in a 10 mM fluoralkylsilane (heptadecafluoro-1,1,2,2-tetrahydrodecyl triethoxysilane) toluene solution for 30 minutes, and then heat at 150°C for 1 hour;

[0011] S4. The deposition of PDMS on the fluorinated wafer is spin-coated at a speed of 1000 rpm for 40 seconds, and cured at 60°C for 4 hours;

[0012] S5. Depositing a conductive layer Cr / Ag / Au on the PDM substrate by sputtering technology;

[0013] S6. Before sputtering the gold layer, a 5nm chromium layer is first deposited on the PDMS to increase the adhesion between the PDMS and the gold layer, so as to achieve better contact;

[0014] S7. A silver layer with a thickness of 400nm is sputtered using a sputtering machine;

[0015] S8. A 20nm gold layer is deposited on the silver strip for oxidation prevention;

[0016] S9. A layer of PDMS is deposited on the gold layer as an insulating layer for preventing the gold connected to the device from collecting any signals from the skin surface.

[0017] An ion-conducting gel is provided, comprising:

[0018] a polymeric binding material, accounting for 5 to 20wt% of the gel; a metal salt, accounting for 2 to 10wt% of the gel;

[0019] a hydrophilic polymeric material, having an average molecular weight of 200 to 15000 Dalton;

[0020] wherein, the polymeric binding material is selected from one or more of polysaccharides, cellulose hydrogels, protein hydrogels, nucleic acid gels, biological hydrogels, and in particular, poly(N-vinylcaprolactam) (PVCL), polyvinylpyrrolidone (PVP), polydopamine, and polyacrylic acid; the hydrophilic polymeric material is a polyol;

[0021] a gel matrix is formed between the polymeric binding material and the hydrophilic polymeric material, and the metal salt is distributed throughout the gel matrix.

[0022] Further, the hydrophilic polymeric material is polyethylene glycol, having an average molecular weight of 200 Dalton.

[0023] Further, the metal salt is one or more of sodium salt, potassium salt, calcium salt, and magnesium salt.

[0024] Further, the metal salt is sodium chloride.

[0025] An electrode with ultra-low bioelectronic impedance, the electrode comprising a first surface and a second surface, the first surface being a substrate, being a metal electrode pad; the second surface being a flexible base material, the metal electrode pad being covered with an ion-conducting gel.

[0026] Further, the first surface comprises one or more metal electrode pads, wherein each of the two or more metal electrode pads is connected with an electrically conductive path, and the two or more metal electrode pads are covered with an ion-conducting gel; each electrically conductive path is covered with an insulating material.

[0027] The beneficial effects of the present application are:

[0028] (1) Improving signal quality, effectively coupling electrodes to the skin, optimizing control accuracy of the prosthesis when amputees use myoelectric prostheses;

[0029] (2) Reducing physiological conflicts and improving the comfort of the prosthesis, enabling amputees to more accurately control the prosthesis, which is conducive to improving their use experience and quality of life. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figures la-b , Figure lc is a schematic diagram of the capacitive coupling of the surface interface and (b) the internal interface;

[0031] Figure 2 is a molecular dynamics (MD) simulation diagram of the interaction of the gel electrode with the stratum corneum;

[0032] Figure 3 is a schematic diagram of measuring the impedance of the electrode on the skin;

[0033] Figure 4 is a schematic diagram of the system bioimpedance of commercial electrodes and PVCL electrodes on the skin of amputees from 1 Hz to 105 Hz;

[0034] Figure 5 is a diagram of the molecular structure of the lipid matrix;

[0035] Figure 6 is a schematic diagram of the PMF and ΔG of the interaction of the five monomers with the SC;

[0036] Figure 7 is a diagram of the molecular structure evolution of the five monomers in the anchoring process;

[0037] Figure 8 is a schematic diagram of the interaction of the five monomers with the SC over time;

[0038] Figure 9 is a schematic diagram of the dynamic penetration process of the five gels to the bio-simulated skin;

[0039] Figure 10 is a histogram of the order parameter (Sz) of the lipid matrix at 310K, 340K and 360K for different monomers;

[0040] Figure 11 is a confocal microscope image of PVP, acrylic acid and DOPA ion conductive gel anchored on the SC matrix;

[0041] Figures 12a-d , Figures 12e-f is a schematic diagram of the electrical properties of the five gel electrodes;

[0042] Figure 13 Schematic of 90° peel test for PVCL, PVP, alginate, acrylic acid and DOPA on a biomimetic skin;

[0043] Figure 14a , Figure 14b Schematic of 180° peel test for PVCL, PVP, alginate, acrylic acid and DOPA, respectively;

[0044] Figure 15a , Figure 15b Schematic of electrode impedance based on five molecular anchors;

[0045] Figure 16a , Figures 16b-f Schematic of myoelectric control;

[0046] Figure 17a , Figure 17b Schematic of myoelectric mapping of amputee finger and wrist movement for G1-G3, G4-G6, respectively;

[0047] Figure 18a , Figure 18b Schematic of unit data sets for G1 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training, respectively;

[0048] Figure 19a , Figure 19b Schematic of unit data sets for G2 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training, respectively;

[0049] Figure 20a , Figure 20b Schematic of unit data sets for G3 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training, respectively;

[0050] Figure 21a , Figure 21b Schematic of unit data sets for G4 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training, respectively;

[0051] Figure 22a , Figure 22b Schematic of unit data sets for G5 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training, respectively;

[0052] Figure 23a , Figure 23bFigures showing the schematic of the unit data set for (a) high-level (>20% MVC) and (b) low-level (<20% MVC) G6 for training, respectively. DETAILED DESCRIPTION

[0053] The technical solutions of the present application will be described below in conjunction with embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0054] Provided is an ion conductive gel, comprising:

[0055] A polymeric binding material accounts for 5 to 20 wt% of the gel; a metal salt accounts for 2 to 10 wt% of the gel;

[0056] A hydrophilic polymeric material has an average molecular weight of 200 to 15,000 Dalton;

[0057] Among them, the polymeric binding material is selected from one or more of polysaccharides, cellulose hydrogels, protein hydrogels, nucleic acid gels, biological hydrogels, and especially poly(N-vinyl caprolactam) (PVCL), polyvinylpyrrolidone (PVP), polydopamine, and polyacrylic acid; the hydrophilic polymeric material is a polyol; a gel matrix is formed between the polymeric binding material and the hydrophilic polymeric material, and the metal salt is distributed throughout the gel matrix; the hydrophilic polymeric material is polyethylene glycol, with an average molecular weight of 200 Dalton; the metal salt is one or more of sodium salt, potassium salt, calcium salt, and magnesium salt; and the metal salt is sodium chloride.

[0058] Specifically, an ion conductive gel for molecular anchoring, comprising:

[0059] (a) 5% sodium chloride to improve the preparation of ion conductive electrodes;

[0060] (b) dissolving PVCL, PVP, dopamine, and acrylic acid in 10% PEG (200) by weight percentage;

[0061] An alginate gel is prepared by mixing water with sodium alginate, ionic crosslinking agent, covalent crosslinking agent, acrylamide, and thermal initiator.

[0062] Provided is a method for preparing an electrode with ultra-low bioelectronic impedance, comprising:

[0063] (a) To make a PDMS substrate, liquid PDMS can be mixed with PDMS curing agent at a ratio of 10:1. Then the mixture is stirred to uniformity using a centrifugal mixer;

[0064] (b) After the mixture is homogenized, it is placed in a vacuum chamber to eliminate air bubbles. Then a layer of PDMS is deposited on the fluorinated silicon wafer;

[0065] (c) The method of making the fluorinated silicon wafer is to immerse the silicon wafer in a 10 mM solution of fluoroalkylsilane (heptadecafluoro-1,1,2,2-tetrahydrodecyltrichlorosilane) in toluene for 30 minutes, then heat at 150 °C for 1 hour;

[0066] (d) The deposition of PDMS is done by spin coating at 1000 rpm for 40 seconds on the fluorinated wafer. The last step is to cure at 60 °C for 4 hours;

[0067] (e) The conductive layer Cr / Ag / Au is deposited on the PDM substrate by sputtering technique;

[0068] (f) Before sputtering the gold layer, a 5 nm layer of chromium is deposited on the PDMS to increase the adhesion between the PDMS and the gold layer, thus achieving better contact;

[0069] (g) A silver layer with a thickness of 400 nm is sputtered using a sputtering machine;

[0070] (h) A 20 nm layer of gold is deposited on the silver strip to prevent oxidation;

[0071] (i) On top of the gold layer, a layer of PDMS is deposited as an insulating layer to prevent the gold connected to the device from collecting any signal from the skin surface.

[0072] Example 1:

[0073] Molecular anchoring of poly(N-vinylcaprolactam) (PVCL) gel electrodes to the skin for ultra-low myoelectric signal monitoring:

[0074] Figure 1 shows schematic diagrams of the capacitive coupling of (a) the surface interface and (b) the internal interface. For a surface without anchoring, the induced charge on the metal layer depends on the transport efficiency of EDL-1 (the interface of the metal layer and the ionically conductive gel layer) and EDL-2 (the interface of the gel and the stratum corneum). With anchoring, the induced charge on the metal layer increases. (c) shows that the impedance of the PVCL electrode anchored on the SC is 93% lower than that of the alginate polyacrylamide (Alg-PAAm) electrode without anchoring. The error in (c) is the standard deviation of 10 independent experiments.

[0075] Figure 2Molecular dynamics (MD) simulations of the gel electrode interaction with the stratum corneum are described. (a) Schematic showing the ionically conductive gel electrode embedded in the uppermost layer of the skin - the stratum corneum (left). Simulations of the stratum corneum (modeled using free fatty acids (FFA), long-chain ceramides (CER), cholesterol (CHOL) as a lipid bilayer) and the ion gel electrode interface (middle) show that the hydrophobic molecules interact closely with the lipid matrix (right). Due to the molecular anchoring of the gel, the coupling area is increased and the coupling efficiency is higher. (b) Potential of mean force (PMF) as a function of penetration depth. Here, we define z=0 as the center of the lipid matrix. From the MD results, it is seen that an energy barrier, AG, needs to be overcome when the anchoring molecules anchor to the lipid matrix. And this AG depends on the hydrophobicity of the molecule. (c) Molecular simulations show that poly(N-vinylcaprolactam) (PVCL) interacts most closely with the lipid matrix among the five monomers (PVCL, polyvinylpyrrolidone (PVP), acrylic acid, dopamine (DOPA), and alginate). To speed up the structural evolution, simulations were performed at 310 K, 340 K, and 360 K using the V-rescale algorithm.

[0076] As Figure 2 As shown in Figure la and lb, the hydrophobic poly(N-vinylcaprolactam) (PVCL) gel electrode can be firmly fixed on the skin to reduce the barrier effect and increase the ion-electron coupling area.

[0077] 1. Electrode preparation method: PDMS substrate preparation:

[0078] To make the PDMS substrate, liquid PDMS and PDMS curing agent are mixed in a ratio of 10:1, and then the mixture is stirred to uniformity using a centrifugal mixer. After mixing, it is placed in a vacuum chamber to eliminate air bubbles, and then a layer of PDMS is deposited on a fluorinated silicon wafer. The fluorinated silicon wafer is made by immersing a silicon wafer in a 10 mM solution of fluorinated alkylsilane (heptadecafluoro-1,1,2,2-tetrahydrodecyltriethoxysilane) in toluene for 30 minutes, and then heat treating at 150 °C for 1 hour. The purpose of fluorination is to form a hydrophobic surface. The deposition of PDMS is done by spin coating on the fluorinated wafer at a speed of 1000 rpm for 40 seconds, and the final step is to cure at 60 °C for 4 hours.

[0079] 2. Electrode preparation method: deposition of chromium / silver / gold and insulator:

[0080] The conductive layer Cr / Ag / Au is deposited on the PDMS substrate by sputtering technology, requiring a very thin layer (nanoscale) so that it still maintains its conformable conductive ability when stretched.

[0081] A 5 nm thick chromium layer was deposited on the PDMS before sputtering the gold layer to increase the adhesion between the PDMS and the gold layer, thus achieving better contact. Then, a 400 nm thick silver layer was sputtered using a sputtering machine. Finally, a 20 nm thick gold layer was deposited on the silver strip to prevent oxidation. A mask was used during the sputtering process to control the deposition of gold in the desired area and to pattern it. Subsequently, a layer of PDMS was deposited on the gold layer as an insulating layer to prevent the gold connected to the device from collecting any signals from the skin surface.

[0082] 3. Electrode preparation method: making ion conductive gel:

[0083] In all gels, 5% NaCl was added to improve ion conductivity. PVCL, PVP, dopamine, and acrylic acid were dissolved in 10% polyethylene glycol (PEG) 200 by weight. The alginate gel was completed in one step by mixing water with sodium alginate, ionic crosslinker, covalent crosslinker, acrylamide, and thermal initiator.

[0084] After the PVCL, PVP, dopamine, and acrylic acid polymer materials were dissolved in PEG 200, the gel was formed.

[0085] 4. System impedance of electrode and skin:

[0086] System impedance was measured by Zahner scientific instrument. During the test, two electrodes with a size of 1 x 1 cm were attached to the skin with a distance of about 1 cm. The frequency range was 1 Hz to 10 5 Hz.

[0087] It can be concluded that the increase in the coupling area leads to an increase in the coupling probability between ion flux and electronic current, thus improving the ion-electron coupling efficiency.

[0088] Refer to Figures 3-4 , Figure 3 for the schematic diagram of measuring the impedance of the electrode on the skin. The distance between the two electrodes (edges) is 1 cm, and the size of the electrode is 0.5 x 0.5 cm. Figure 4 The system bioimpedance of commercial electrodes and PVCL electrodes on amputee skin from 1 Hz to 105 Hz is shown. The impedance of the PVCL electrode at 1 Hz is almost 1 / 100 of that of the commercial acrylic electrode.

[0089] The measurement on amputee skin (see Figure 3 ) shows that the impedance of the PVCL electrode is about 20 kΩ@1 Hz, almost 1 / 100 of the impedance of the commercial acrylic-based electrode (see Figure 4 ), which is 93.4% lower than that of the alginate polyacrylamide (Alg-PAAm) electrode, which shows the lowest bioelectronic impedance on normal skin (seeFigure lc This indicates that PVCL gel electrodes can be used to monitor ultra-low (as low as 1.5% MVC) electromyographic signals in amputees under dexterous prosthetic control.

[0090] Example 2. Molecular dynamics (MD) simulation of the molecular anchoring process of polyvinyl chloride, polyvinylpyrrolidone (PVP), acrylic acid, dopamine (DOPA), and alginate:

[0091] Molecular dynamics (MD) simulations were performed to understand how PVCL molecules interact with and anchor to SC. Four other adhesive molecules—polyvinylpyrrolidone (PVP), acrylic acid, dopamine (DOPA), and alginate—were investigated to explain the molecular anchoring strategy.

[0092] 1. Single-unit structure modeling:

[0093] The monomers of five classes of molecules—PVCL, PVP, acrylic acid, dopamine (DOPA), and alginate gel—were studied, as shown in Figure 1b. The structures of these monomers were obtained through the CHARMM-GUI web service (https: / / www.charmm-gui.org / ).

[0094] 2. Skin surface modeling:

[0095] According to previous studies, the skin surface, namely the stratum corneum (SC), is modeled using the lipid matrix of the stratum corneum.

[0096] like Figure 2 a and Figure 5 As shown, the lipid matrix of SC was constructed using a 1:1:1 mixture of long-chain ceramides (CER), cholesterol (CHOL), and free fatty acids (FFA) via the CHARMM-GUI web service. The lipid matrix size was approximately 10. 10 To determine the interaction between the monomer and SC, such as Figure 5 As shown, Figure 5 The molecular structure of the lipid matrix is ​​described. (a) Computer-generated structures of lipid ceramide (CER), cholesterol (CHOL), and free fatty acids (FFA). (b) Simulation showing monomer molecules anchored to the lipid matrix. 700 monomers were randomly added to the simulation frame.

[0097] 3. Molecular dynamics simulation methods:

[0098] Molecular dynamics (MD) simulations were performed by the GROMACS software package and CHARMM36 force field. TIP3P water molecules were used to solvate the system, and an appropriate amount of Na+ions were added to neutralize the system. Simulations were performed under periodic boundary conditions. All graphics and visual analysis were completed by the Visual Molecular Dynamics VMD software package.

[0099] In line with the inherent hydrophobicity of the lipid matrix (composed of FFA, CER, and CHOL), the calculated potential of mean force (PMF) shows that the more hydrophobic monomers have a lower energy barrier (AG) to anchor to the lipid matrix and are more easily embedded (see Figure 2 b).

[0100] Referring to Figure 7 , the molecular structural evolution of the five monomers during the anchoring process is described. Simulations performed at high temperatures (310 K to 360 K) accelerate the interaction of monomers with lipids. Among the five monomers, PVCL has the deepest degree of anchoring to the lipid matrix.

[0101] It can be concluded that among the five monomers, PVCL has the highest hydrophobicity, and therefore the lowest AG value to anchor to SC (6.57 k B T). PVCL and PVP are widely used in biological and medical fields and have good biocompatibility and chemical stability. Although PVP and PVCL have high hydrophobicity and homologous structure, the additional methylene group in PVCL makes it more hydrophobic than PVP. Therefore, the octanol-water partition coefficient (logP) of PVCL is 1.75, which is greater than that of PVP (0.74) (see Figure 6 ). Figure 6 The PMF and AG of the five monomers when interacting with SC are described: (a) PVCL, (b) PVP, (c) acrylic acid, (d) DOPA, and (e) alginate monomer PMF calculation results. (f) Among the five monomers, PVCL has the lowest energy barrier (6.57 kBT) when anchoring in the lipid matrix. The energy barrier of alginate monomer is the highest ( =40.67 kBT), and it is not easy to anchor on the lipid matrix. Acrylic acid, DOPA, and alginate are suitable adhesives and are widely used in electromyographic electrodes. The logP values of acrylic acid (0.44), DOPA (-2.2), and alginate (-3.84) are much lower than those of PVP and PVCL. The alginate with the least hydrophobicity shows the highest energy barrier ( =40.67 k B T), and it is the least able to anchor to the lipid matrix.

[0102] Example 3. Molecular dynamics (MD) simulations were performed at different temperatures with the V-rescale algorithm to accelerate structural evolution:

[0103] In addition, to accelerate the structural evolution, MD simulations were performed at three temperatures, 310 K, 340 K and 360 K, respectively, using the V-rescale algorithm to investigate the interactions between five monomers and the lipid matrix.

[0104] 1. Coupling of molecular dynamics (MD) simulations with the V-rescale algorithm:

[0105] The MD simulations were performed according to the protocol in Example 2.

[0106] To accelerate the structural evolution, the V-rescale algorithm (Bussi, G. et al., J. Chem. Phys., 2007, 126, 014101) was used in the simulations, which were performed at three temperatures, 310 K, 340 K and 360 K, respectively. The system was first equilibrated in an NPT ensemble, in which the pressure was coupled using the Parrinello-Rahman method (Parrinello, M. & Rahman, A., J. Appl. Phys., 1981, 52, 7182) under semi-isotropic conditions at 1 bar. In the equilibration simulations, position restraints were applied to all heavy atoms of the monomers and the lipid matrix. After about 1 ns of equilibration simulation, all production simulations were performed in an NVT ensemble. The LINCS algorithm was used to constrain the covalent bonds of hydrogen atoms, the time step was set to 2 fs, the cutoff for non-bonded interactions was set to 1.2 nm, and the long-range electrostatic interactions were calculated using the particle mesh Ewald (PME) method.

[0107] 2. Umbrella sampling method:

[0108] To calculate the average force potential (PMF) of different monomers penetrating the lipid matrix of the skin, the umbrella sampling method was used. In the umbrella sampling method, the reaction coordinate was defined as the distance between the center of mass (COM) of the monomer and the lipid matrix in the normal direction, as shown in Figure 5 To construct the system structure at different reaction coordinates, the PLUMED software package was used to add monomers to the corresponding reaction coordinates. In the umbrella sampling simulation, the size of the lipid matrix was about 5.4 x 5.4 nm 2 The sampling window interval was 0.1 nm. There were 41 windows in total for each system. To ensure convergence, each window was subjected to 20 ns of MD simulation, so the total simulation time for each monomer in the umbrella sampling was 800 ns. The spring constant for position restraint in the umbrella sampling method was 1660.5 The weighted histogram analysis method (WHAM) was used to analyze the umbrella sampling results.

[0109] 3. Calculation of the order parameter of the lipid matrix:

[0110] To describe how the monomers affect the lipid matrix structure, the order parameter S of the lipid matrix was calculated z The following equation:

[0111] ;

[0112] Where, is the angle between the vector (average direction of the lipid matrix) and (the average direction of the lipid matrix) and (the vector from to the vector on the alkyl chain). The bracket indicates the average value of the carbon atoms in the lipid alkyl chain.

[0113] In general, S z = 1 indicates that the lipids in the matrix are completely aligned with the normal direction; S z = 0 indicates that the lipids in the matrix are randomly aligned; S z = 1 / 2 indicates that the lipids are aligned perpendicular to the normal direction of the matrix.

[0114] It can be concluded that the results of further MD simulation using the V-rescale algorithm at three temperatures (310K, 340K and 360K) to accelerate the interaction also show that PVCL can be embedded in SC most effectively and fastest (see Figure 2 c、 Figure 8 and Figure 9 ). Figure 2 c shows the interaction between the five molecules and the lipid matrix between 310K and 360K from 0ns to 100ns. The strength and depth of the interaction increases with the increase of the hydrophobicity of the monomer. Figure 8 The interaction of the five monomers with SC over time is described. At 0 nanoseconds, there is no interaction between the monomers and SC. Within 100 nanoseconds, all monomers except alginic acid gradually penetrate into SC, with PVCL penetrating the fastest. The simulation was carried out at 360K. Figure 9Five dynamic permeation processes of gels into bio-skin are described. (a)-(c) Details of the order parameter, Sz, of the lipid matrix as a function of time at (a) 310 K, (b) 340 K, and (c) 360 K. (a) The inset describes the selection method of the angle, Θ, that determines the value of Sz. Apparently, at 310 K, the Sz of all gels and the control group is around 0.65, which means the permeation process is not fast. When the temperature is raised to 340 K, the Sz of PVCL gradually changes to 0 as the time increases to 100 ns, which indicates that PVCL interacts with the SC matrix more strongly than the other 4 monomers. Therefore, alginate is difficult to permeate into the SC. (d)-(e) Confocal microscopy images show that PVCL is anchored at a depth of 3.07 pm on the bio-skin (d), while alginate is not anchored (e). The bio-skin is stained with rhodamine 6G, and the gels are stained with Alexa 488. (f) Quantitative measurements show that among the 5 monomers, PVCL is the deepest anchored on the bio-skin. The error is the standard deviation based on 10 independent samples.

[0115] The order parameter, S, of the lipid matrix was calculated z to describe the effect of monomers on the structure of the lipid matrix. The ordered lipid matrix without monomers (control group) has a higher degree of order within the 100 ns simulation time scale at 310 K (see Figure 9 a). The order parameter, S, of the ordered lipid matrix without monomers (control group) is around 0.65 (see a) (see Figure 9 a) (see Figure 10 the control simulation without monomers. Figure 10 Histograms of the order parameter, Sz, of the lipid matrix at 310 K, 340 K, and 360 K are shown for different monomers. Consistent with the PMF calculation in Figure 6 , the Sz of PVCL is the lowest at 340 K and 360 K, indicating that it has the largest disruption to the lipid structure. The error bars are the standard deviations of the Sz values in the last 20 ns of the simulation.

[0116] Since the molecular anchoring process is a long time scale process, the order parameter of the lipid matrix remains around 0.65 in the 100 ns simulation at 310 K, similar to the order parameter without the anchoring process. When interacting with the PVCL monomer at 340 K, it drops to 0.072 (see Figure 9 b), indicating that the monomer greatly disrupts the order of the lipid matrix (see Figure 2 c). For all other monomers, the remains around 0.55 at 340 K. When the system is heated to 360 K, these monomers have different effects on the structure of the lipid matrix (see Figure 9 (c) andFigure 10 ). As shown in Figure 10 Figure 6, the simulation concluded that PVCL has the strongest anchoring ability on the skin among the five molecules.

[0117] Example 4. Penetration process of gel electrodes on biomimetic skin:

[0118] 1. Penetration test:

[0119] First, the biomimetic skin was immersed in rhodamine 6G (0.002 mg / mL, solvent: ethanol) for 72 hours. Then, the biomimetic skin was rinsed with ethanol for 3 times and dried. Second, the ionically conductive gel was mixed with Alexa 488 (0.004 mg / mL, solvent: dimethyl sulfoxide (DMSO)) for 24 hours. Third, the dyed gel was transferred to the biomimetic skin with rhodamine 6G. The penetration process was observed using a Leica confocal microscope.

[0120] It can be concluded that the confocal microscope images also showed that the anchoring depth of PVCL on the biomimetic skin made of cellulose was 3.07 pm, which was deeper than PVP (2.43 pm), acrylic acid (1.18 pm), DOPA (0.97 pm) and alginate (0 pm) (see Figure 9 d-f and Figure 11 ). Figure 11 Confocal microscope images of PVP, acrylic acid and DOPA ionically conductive gels anchoring on the SC matrix are described. The dotted line represents the depth of the gel embedded in the SC (fibrin-based biomimetic skin). All gels, except alginate, anchored on the SC matrix. The anchoring depth of PVCL was the deepest, 3 pm.

[0121] The results showed that among the 5 monomers, the PVCL monomer with the strongest hydrophobicity had the best anchoring effect with the skin.

[0122] Example 5. Preparation of gel electrodes and evaluation of mechanical and electrical properties:

[0123] According to the simulation results, five gel electrodes were fabricated with these monomers and their mechanical and electrical properties were evaluated (see Figure 12). Figure 12 depicts the electrical properties of the five gel electrodes. (a) Photograph (left) and schematic (right) of a gel electrode consisting of a polydimethylsiloxane (PDMS) substrate and an insulator and chromium / silver / gold metal contacts. (b) Adhesion force of different electrodes obtained using a standard 90° peel test with a biomimetic skin. Consistent with the simulation study, the PVCL electrode showed the highest adhesion force (about 189 N / m). (c) Graph showing the impedance of various electrodes on human skin from 1 Hz to 105 Hz. (d) Comparison of the impedance and adhesion force of our PVCL electrode with different electrodes: 1 : CNT / aPDMS, 2: PDMS_40 NW, 3: PDMS_40NW / Tape, 4: Gold / poly-para-xylylene, 5: Wire, 6: Commercial electrode, 7: Wire / gold, 8: Fe@Sibione, 9: PDA-rGO-PAM, 10: a4-PDMS_40 NW, 11 : Alg-PAAm. The impedance decreases with the increase of adhesive electrodes (curves). The shaded area is the intrinsic impedance of human skin. The system impedance (total impedance of skin, electrode, and interface) of the PVCL electrode on human skin (14 kQ) is close to the impedance of human skin (6-10 kQ). (e) The recording limit of the PVCL electrode on human skin (about 1.5% MVC) is close to the lower limit of human muscle contraction (about 1% MVC). Signal loss occurs when the muscle contraction is below 10% MVC for DOPA, acrylic acid, and alginate electrodes (shaded area) (all data were recorded under the same 24x magnification, Y-axis is relative intensity, curves are offset by about 5 mV). (f) The signal-to-noise ratio of the PVCL electrode is the best among the 5 electrodes (> 5) when recording low-level (1.7% MVC) electromyographic signals on human skin. A signal-to-noise ratio greater than 5 is required for traditional recognition systems. Error bars are based on the standard deviation of 10 independent samples (all data were recorded under the same 24x magnification conditions).

[0124] 1. Fabrication of gel electrodes:

[0125] Gel electrodes were prepared according to the protocol in Example 1. PDMS was used as the substrate and insulator for all electrodes. To improve ionic conductivity, 5% NaCl was added to all gels.

[0126] 2. Signal-to-noise ratio equation (SNR), formula:

[0127] SNR = 10 log10(P signal / P noise );

[0128] where P signal is the average of pixel values; P noiseis the standard deviation or error value of the pixel values.

[0129] It can be concluded that Fig. 12a shows the electrode structure, PDMS is used as the base and insulator of the electrode to fit the curved amputee skin well.

[0130] The adhesion between the electrode and the skin is the macroscopic evidence of the molecular anchoring (see Figure 2 (b)), which is due to the strong interaction between the gel and the SC matrix. In line with the simulation results in Figure 6 f, the standard 90° peel test performed on the fibrin-made bio-skin showed that the adhesion of PVCL was the strongest (about 189 N / m; Fig. 12b and Figure 13 ). Figure 13 The 90° peel test of PVCL, PVP, alginate, acrylic acid and DOPA on the biomimetic skin is described, showing that the adhesion of PVCL is superior to other materials. Error bars are standard deviations based on 10 independent samples. Similarly, in the 180° peel test, the PVCL gel is still superior to other gels (see Fig. 14). Fig. 14 describes the 180° peel test of PVCL, PVP, alginate, acrylic acid and DOPA. Again, among the five gels, PVCL has the strongest adhesion. This result is in line with the 90° peel test results shown in Figure 9 b and Figure 13 . Although the intrinsic impedance of all five electrodes at 1 Hz is between 00-300 Ω (see Figure 15a ), the impedance of the PVCL electrode at 1 Hz is the lowest for normal skin (about 14 kΩ) and amputee skin (about 20 kΩ) (see Fig. 12c, Figure lc and Figure 15b ). Fig. 15 describes the impedance of the five molecular anchor-based electrodes. (a) The intrinsic impedance of the five ionically conductive gels from 1 to 104 Hz. (b) The system bioimpedance of the five electrodes. Error bars are standard deviations of 20 gel electrode samples.

[0131] These values are closer to the intrinsic impedance of human skin (6-10 kΩ) than previous electrodes, and low-level (1.5% MVC) myoelectric signals close to the human limit (about 1% MVC) can be detected (see Fig. 12d-e). The key reason is that the EDL formed by the molecular anchoring process has a very high ion-electron coupling efficiency. The impedance of the original gel without NaCl is 107 Ω at 1 Hz. These results show that the fixation of PVCL on the SC can improve ion-electron coupling and significantly reduce the skin impedance.

[0132] In the other four electrodes, the super-low intensity muscle contraction (less than 10% MVC) was monitored, and serious signal loss always occurred due to the high bio-impedance. As an important parameter to describe the fidelity of the electrode, the signal-to-noise ratio (SNR) of the recorded myoelectric signal less than 30% MVC was measured (see Fig. 12(f)). Except for the SNR of the 1.7% MVC signal of the PVCL electrode 1, which was 5.02, and the SNR of the 1.5% MVC signal, which was 3.725, the SNR of all electrodes less than 2% MVC was less than 5. The traditional recognition algorithm based on raw myoelectric signal, such as linear discriminant analysis (LDA), cannot ensure high recognition accuracy for the weak activity of the phantom limb of amputees. Since the signal source of myoelectric control requires a signal-to-noise ratio of more than 5, the PVCL electrode has the potential to detect super-low myoelectric signals and thus achieve fine prosthetic control.

[0133] Example 6. Image recognition algorithm based on convolutional neural network (CNN):

[0134] Convolutional neural network (CNN) is a network first proved to be effective for recognizing handwritten graphics on MNIST dataset (http: / / yann.lecun.com / exdb / mnist / ) by Lecun. Convolutional network is considered to have the ability to extract spatial features of images.

[0135] 1. Convolutional neural network (CNN):

[0136] In order to use convolutional neural network (CNN) for image recognition, the myoelectric signal is converted into an image by 1 x n x m pixel mapping, where n is the number of channels on the electrode array, and m is the time required for each muscle contraction in the expected movement of the finger and wrist (in work, m is 200 ms). The size of the electrode array depends on the size of the monitored muscle (the maximum data processing capacity of the system, n max is 128). As a training signal source, the myoelectric signal of low-level (≤20% MVC), high-level (≥20% MVC) and maximum level (100% MVC) muscle contraction is defined.

[0137] By designing the Convnet model, the basic and complex features of the myoelectric mapping can be detected and learned by convolution operation. The basic Convnet includes several layers: conv2D layer, maxpooling layer, dense layer, dropout layer, batchnormalization layer and softmax layer.

[0138] The ReLU function adopted contains the nonlinearity of the network. Adam was chosen as the network optimizer. The structure of the Convnet is shown in Figure 16b. Figure 16 describes dexterous myoelectric control. (a) Photograph of our myoelectric prosthetic control system on a transradial amputee. Gel electrodes attached to the residual active muscles are connected to a wireless data logging box via flexible cables, and the size of the electrode array depends on the size of the muscle being measured. (b) Real-time pixel images of electromyography (EMG) signals obtained from the wrist extensor muscle of an amputee using an n = 16 channel PVCL electrode array (left; size 1 x 16 x 200 ms). The pixel images are input into a convolutional neural network (CNN)-based image recognition architecture (right) that can extract and process features and classify them according to finger and wrist movements of the amputee. (c) The EMG pixel images (left) are processed by the mean average value (MAV) attention block (middle rectangular box) shown in (b) and returned to a feature map (right) that can predict real-time EMG, theoretically reducing the recognition time from 200 ms to 20 ms. (d) Confusion matrix for G1 to G6 finger and wrist movements using low-level (≤20% MVC) muscle contractions obtained using PVCL electrodes shows that the CNN algorithm has a high accuracy (97.6%). The rows represent the actual classification of G1-G6, and the columns represent the predicted classification. The values of the diagonal elements represent the degree of correct prediction of the class, and the reset is the EMG signal without finger and wrist movements. (e) Recognition accuracy of the CNN for low-level signals (≤20 MVC; CNN-L) and high-level signals (≥20 MVC; CNN-H). (f) Photographs show that a transhumeral amputee controlled a robotic car (prosthesis) with his myoelectrically controlled arm to grasp a candle (1.5 cm in diameter) and place it in a hole (2 cm in diameter). The control signals were derived from real-time EMG maps generated from G1-G6. For the conv3d layer, the number of filters for the first conv2D layer was set to 16, and the kernel size was set to 4 x 2; the number of filters for the second conv2D layer was set to 32, and the kernel size was set to 3 x 3, and the third kernel size was set to 2 x 2. The first dense layer had 64 units, and the second dense layer had 10 units, corresponding to the number of classified movements.

[0139] It can be concluded that the CNN-based electromyography spectrum recognition includes several layers, which respectively extract features from the electromyography spectrum (conv2D layer), reduce model parameters (maxpooling layer), accelerate network convergence (batch normalization layer), prevent overfitting (dropout layer), and output prediction tensor (dense and softmax layer) (see Figure 16b). The conv2d layer is mainly used for convolution operation to extract features from the electromyography spectrum. The maxpooling layer performs downsampling operation to reduce model parameters. The batch normalization layer prevents gradient disappearance and makes the network converge faster. The Dropout layer randomly disables neural cells, solving the overfitting problem. Through the dense layer and the soft maximum layer, the Convnet finally outputs the predicted tensor, the number of which is equal to the number of classes.

[0140] Example 7. Introducing attention mechanism on electromyography spectrum:

[0141] In order to improve the processing efficiency and classification accuracy of ultra-low electromyography signals with low signal-to-noise ratio (<5), an attention block (MAV block) is designed, which contributes to the network in two ways: (1) evaluating the importance of each channel, enhancing channels with obvious features, while suppressing channels with no obvious features; (2) extracting MAV features from different channel electromyography signals and retaining these features for transfer learning, which can accurately complete new tasks in a short training time (see Figure 16c). Since the same muscles are used in different finger and wrist movements, the attention block evaluates and weighs each muscle channel by measuring the mean (MAV). This block increases the weight of important channels and reduces the weight of unimportant channels.

[0142] 1. Attention block:

[0143] In the general task of attention blocking, people usually pay attention to the similarities and differences between the source domain and the target domain. Similar features should be retained, while different features should be learned and calibrated. Even a slight change in muscle contraction strength in the task can have a huge impact on gesture classification. Since electromyography samples with different contraction strengths are independent, it is difficult for the CNN network to learn the obvious different offset reasons, so the network needs to pay more attention to the similar features between different movement force samples, and use fine-tuning to calibrate the network.

[0144] Assume that the electromyography signal has n channels, each with m sampling points. It can be regarded as an input image, the formula is:

[0145] ;

[0146] For the channel, first calculate the mean (MAV), the formula is:

[0147] ;

[0148] Then, the MAV raw value is obtained according to the formula:

[0149] ;

[0150] By passing through 2 MLP layers and activating with sigmoid function, 8 weighted vectors in the following formula are obtained:

[0151] ;

[0152] wherein, represents the sigmoid function;

[0153] Finally, the output image is represented by the formula:

[0154] ;

[0155] wherein, F scale represents column multiplication.

[0156] It can be concluded that attention mechanisms have been widely used in deep learning to improve network performance and achieve higher classification accuracy. An attention block is designed, named MAV block. The structure of MAV block is shown in FIG. 16c.

[0157] After extracting the same features from 100% MVC electromyography, a series of activated electrodes and related weights for each finger and wrist movement were obtained. The 100% MVC map can exclude the influence of signals with low signal-to-noise ratio. The output after using the MAV block (called feature map) is used as the initial judgment of real-time electromyography, which can theoretically reduce the delay time in the recognition process from 200 ms to 20 ms.

[0158] Example 8. PVCL-based gel electrode array for measuring low-level myoelectric signals:

[0159] As a proof of concept, PVCL-based gel electrode arrays were fabricated and used to measure low-level myoelectric signals on the residual limb of an amputee. The electrodes were connected to a wireless data recording box through a flexible cable, and the measured myoelectric signals were used to control a prosthesis, which was represented by a robotic car in the work (see FIG. 16). The focus was on how to utilize low-level myoelectric signals to achieve dexterous finger movements, such as grasping with a prosthesis.

[0160] 1. Fabrication of gel electrodes:

[0161] The gel electrodes were prepared according to the protocols in Examples 1 and 5.

[0162] 2. Image recognition with CNN:

[0163] CNN-based electromyographic pattern recognition was performed following the protocol in Example 6. Following the protocol in Example 7, the electromyographic pixel images were placed in the MAV attention block, improving the processing efficiency and classification accuracy of low signal-to-noise ratio (<5) ultra-low electromyographic signals.

[0164] 3. Control of the arm of a below-elbow amputee driving a robotic car with electromyography:

[0165] The system was applied to enable a below-elbow amputee to grasp a candle and place it in a hole. For amputees, it is a common daily activity to use prosthetic fingers to grasp small objects with reasonable force and to approach targets with high precision.

[0166] Figure 18 describes the unit dataset for G1 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training. The time of the unit dataset is 2 seconds. To extract features, the dataset will be divided into 10, each packet of approximately 200 milliseconds.

[0167] Figure 19 shows the unit dataset for G2 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training. The time of the unit dataset is 2 seconds. To extract features, the dataset will be divided into 10, each packet of approximately 200 milliseconds.

[0168] Figure 20 describes the unit dataset for G3 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training. The time of the unit dataset is 2 seconds. To extract features, the dataset will be divided into 10, each packet of approximately 200 milliseconds.

[0169] Figure 21 describes the unit dataset for G4 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training. The time of the unit dataset is 2 seconds. To extract features, the dataset will be divided into 10, each packet of approximately 200 milliseconds.

[0170] Figure 22 describes the unit dataset for G5 (a) high level (> 20% MVC) and (b) low level (< 20% MVC) for training. The time of the unit dataset is 2 seconds. To extract features, the dataset will be divided into 10, each packet of approximately 200 milliseconds.

[0171] Figure 23 depicts the unit dataset for training G6 (a) high level (>20% MVC) and (b) low level (<20% MVC). The time of the unit dataset is 2 seconds. To extract the features, the dataset will be divided into 10, each packet is about 200 milliseconds.

[0172] To show the accuracy of the prosthesis control, the diameter of the hole (2 cm) is only slightly larger than the diameter of the candle (1.5 cm). The six finger and wrist movements (G1~G6) initiated by the amputee (see Figure 17) are used to drive the robotic arm to release (G1, see Figure 18), grip (G2, see Figure 19), move left (G3, see Figure 20), move right (G4, see Figure 21), move up (G5, see Figure 22), and move down (G6, see Figure 23), respectively. Figure 17 depicts the electromyographic mapping of the amputee's finger and wrist movements. The photo of the amputee (left column), with our electromyographic control system installed on the left arm, and the right arm shows the types of finger and wrist movements (G1~G6) initiated by the amputee (red dashed line) as a reference. The pixel image of 100% MVC, >20% MVC (high level), <20% MVC (low level) electromyographic signals and the corresponding feature mapping of each movement (right column) are used to drive the prosthesis (in our study, the arm of the robotic car) to release (G1), grip (G2), move left (G3), move right (G4), move up (G5), and move down (G6). Based on the electromyographic map at 100% MVC, the active electrodes and weights of G1-G6 are preliminarily extracted by attention blocks. The feature mapping is the result of extraction and can be used as a preliminary judgment for real-time electromyographic mapping. In theory, the pre-judgment can shorten the delay time of the recognition process from 200 milliseconds to 20 milliseconds. The high level and low level electromyographic mapping can be used as the source for further fine-tuning to further improve the accuracy of classification. The gray level of the electromyographic mapping is the relative intensity of muscle contraction.

[0173] It can be concluded that among all 5 electrodes, the low-level signal CNN recognition accuracy of the PVCL electrode is the highest (97.6%) (see Figures 16d-e, Table 1 and Table 2), enabling the robotic arm to successfully grasp the candle and place it into the hole (see Figure 16f). Between 22.5 seconds and 36.4 seconds, the robotic arm uses ultra-low level and high-quality electromyographic mapping to skillfully approach the small hole. To enable the robotic arm to grasp the candle, the intensity of muscle contraction of the expected movement generated by the amputated limb must match the intensity of the normal finger movement. If not, it cannot be grasped. In summary, the results of the study show that the PVCL electrode firmly fixed on the skin can detect low-level electromyographic signals suitable for CNN-based recognition and fine prosthesis control.

[0174] Table 1. Confusion matrix for high level muscle contraction (<20% MVC) using CNN

[0175] G1 G2 G3 G4 G5 G6 Reset G1 107 0 0 0 0 0 4 G2 0 96 0 0 0 0 0 G3 0 0 87 0 0 0 0 G4 0 0 0 88 0 0 3 G5 0 0 0 0 75 0 0 G6 0 0 0 0 3 73 0 Reset 0 2 2 2 4 1 95

[0176] Table 2. Confusion matrix for high level muscle contractions (>20% MVC) using CNN

[0177] G1 G2 G3 G4 G5 G6 Reset G1 107 0 0 0 0 0 4 G2 0 96 0 0 0 0 0 G3 0 0 87 0 0 0 0 G4 0 0 0 88 0 0 3 G5 0 0 0 0 65 0 0 G6 0 0 0 0 3 73 0 Reset 0 2 2 2 4 3 95

[0178] A strategy to improve the detection and fidelity of low level myoelectric signals by developing a PVCL gel electrode anchored on SC is introduced. This robust anchoring enhances the iono-electronic coupling and minimizes the impedance across the amputated skin (as low as 20 kQ). The overall mechanical and electrical performance of the electrode outperforms other gel electrodes (Table 3) and enables the detection of ultra-low (1.5% MVC) myoelectric signals close to the human limit. A mapping recognition algorithm based on CNN is designed that can classify and predict the intended finger and wrist movements from the myoelectric signals obtained using the PVCL electrode with an accuracy of up to 97.6%. With such high fidelity signals and classification accuracy, the system enables amputees to dexterously drive the prosthetic hand with their fingers. This fine motor action promises to improve the quality of life of amputees who rely on prosthetics to accomplish their daily tasks.

[0179] Table 3. Mechanical and electrical performance of five gel electrodes

[0180] Alg-PAAm DOPA Acrylic acid PVP PVCL LogP -3.84 -2.2 0.44 0.74 1.75 AG 40.64k B T]] 12.36k B T]] 9.81k B T]] 10.05k B T]] 6.57k B T]] Biological skin anchoring depth 0 pm 0.97 pm 1.18 pm 2.43 pm 3.07 pm Biological electronic impedance of normal skin 26 kQ 431 kQ 217 kQ 16 kQ 14 kQ Adhesion force 90 N / m 30 N / m 45 N / m 151 N / m 190 N / m Detection limit 2.1% / 10.3% 1.7% 1.5% CNN accuracy 89.61% 91.2% 90.05% 90.47% 97.59%

[0181] From the research of the above embodiments, it can be seen that the ionically conductive gel and the PVCL electrode prepared therefrom significantly improve the monitoring ability of low intensity myoelectric signals through its superior electrical performance, biocompatibility and strong adhesion, provide more accurate and flexible prosthetic control schemes for amputees, and thus improve their quality of life.

[0182] The above only describes the preferred embodiments of the present application, and it should be understood that the present application is not limited to the forms disclosed herein, and should not be considered as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concepts described herein, by the above teachings or related art or knowledge. Any modification and change made by those skilled in the art without departing from the spirit and scope of the present application shall be within the protection scope of the claims of the present application.

Claims

1. A method of preparing an electrode having ultra-low bioelectronic impedance, characterized by, Comprising: S1. Making a PDMS substrate, mixing liquid PDMS with PDMS curing agent at a ratio of 10:1, then using a centrifugal stirrer to stir the mixture until uniform; S2. After the mixture is uniform, it is placed in a vacuum chamber to eliminate bubbles, and then a PDMS layer is deposited on a fluorinated silicon wafer; S3. The method for making a fluorinated silicon wafer is to immerse the silicon wafer in a 10 mM fluorinated alkylsilane toluene solution for 30 minutes, then heat at 150°C for 1 hour; S4. The PDMS is deposited on the fluorinated wafer by spin coating at a speed of 1000 rpm for 40 seconds, and then cured at 60°C for 4 hours; S5. Deposit the conductive layer Cr / Ag / Au on the PDM substrate by sputtering technology; S6. Before sputtering the gold layer, deposit a 5nm chromium layer on the PDMS to increase the adhesion between the PDMS and the gold layer, so as to achieve better contact; S7. Use a sputtering machine to sputter a silver layer with a thickness of 400nm; S8. Deposit a 20nm gold layer on the silver strip to prevent oxidation; S9. On the gold layer, deposit a layer of PDMS as an insulating layer to prevent the gold connected to the device from collecting any signals from the skin surface.

2. An electrode having an ultra-low bioelectronic impedance, characterized by, The electrode with ultra-low bioelectronic impedance can be prepared by the method of claim 1, comprising a first surface and a second surface, the first surface is a substrate, which is a metal electrode pad; the second surface is a flexible substrate material, and the metal electrode pad is covered with an ion-conducting gel.

3. An electrode with ultra-low bioelectronic impedance according to claim 2, characterized in that, The first surface comprises one or more metal electrode pads, wherein each of the two or more metal electrode pads is connected to a conductive path, and the two or more metal electrode pads are covered with an ion-conducting gel; each conductive path is covered with an insulating material.

Images