Fabrication of a Micro-Needle Array Electrode by

Download PDF
2 downloads 0 Views 7MB Size Report
Comment
Jun 17, 2016 - showed that the MAE could record EMG, ECG, and EEG signals ..... of the human body always consist of resistance and capacitive resistance.
sensors Article

Fabrication of a Micro-Needle Array Electrode by Thermal Drawing for Bio-Signals Monitoring Lei Ren, Qing Jiang, Keyun Chen, Zhipeng Chen, Chengfeng Pan and Lelun Jiang * Guangdong Provincial Key Laboratory of Sensor Technology and Biomedical Instrument, Sun Yat-Sen University, Guangzhou 510006, China; [email protected] (L.R.); [email protected] (Q.J.); [email protected] (K.C.); [email protected] (Z.C.); [email protected] (C.P.) * Correspondence: [email protected]; Tel.: +86-20-3933-2153 Academic Editor: Vittorio M. N. Passaro Received: 21 April 2016; Accepted: 19 May 2016; Published: 17 June 2016

Abstract: A novel micro-needle array electrode (MAE) fabricated by thermal drawing and coated with Ti/Au film was proposed for bio-signals monitoring. A simple and effective setup was employed to form glassy-state poly (lactic-co-glycolic acid) (PLGA) into a micro-needle array (MA) by the thermal drawing method. The MA was composed of 6 ˆ 6 micro-needles with an average height of about 500 µm. Electrode-skin interface impedance (EII) was recorded as the insertion force was applied on the MAE. The insertion process of the MAE was also simulated by the finite element method. Results showed that MAE could insert into skin with a relatively low compression force and maintain stable contact impedance between the MAE and skin. Bio-signals, including electromyography (EMG), electrocardiography (ECG), and electroencephalograph (EEG) were also collected. Test results showed that the MAE could record EMG, ECG, and EEG signals with good fidelity in shape and amplitude in comparison with the commercial Ag/AgCl electrodes, which proves that MAE is an alternative electrode for bio-signals monitoring. Keywords: micro-needle array; electrode; impedance; EMG; ECG; EEG; finite element method

1. Introduction Bioelectricity is a typical physiological phenomenon of humans which can provide important information for diagnostics and treatment. However, bioelectric signals are weak and difficult to collect. Electrodes, including wet electrodes (typical Ag/AgCl wet electrode) and dry electrodes, can collect and record bioelectric signals [1]. The micro-needle array electrode, as a dry electrode, has attracted more and more attention for physiological electrical signals monitoring (including EII, ECG, EEG, and EMG recording) in recent decades [2]. In the EII test, the MAE presents less variation of contact impedance and better stability due to the significant improvement of the contact interface between the electrode and skin, compared with conventional Ag/AgCl electrodes and metal planar bio-electrodes [3,4]. In an EMG signals test, both the MAE and Ag/AgCl electrode can trace the change of EMG signals well and easily reduce the crosstalk [2,5]. In static state recording of ECG signals, signals collected by the MAE agree well with that recorded by typical Ag/AgCl electrodes [2,5–9], while in a dynamic state recording of ECG signals, the MAE can capture more obvious ECG principal features due to the reduction of motion artifacts and a stable contact interface between the MAE and human skin [2,7]. In EEG signals measurement, the MAE can pass through the hairs, reach the scalp, and penetrate through the dead skin layer without any obstruction. Sensing performance of the MAE is better in terms of electrode-skin impedance over a long period of recording, as well as a much higher efficiency in preparation of EEG measurements compared with typical Ag/AgCl electrodes [2,10,11]. Furthermore, Ag/AgCl electrodes work well in short term monitoring but, subsequently, the gel dries, causing disruption in the signals [8]. The electrode gel may also cause itching, allergic reactions, Sensors 2016, 16, 908; doi:10.3390/s16060908

www.mdpi.com/journal/sensors

Sensors 2016, 16, 908

2 of 13

and other skin problems in some patients [2,12]. The MAE can be used without gel and can directly pierce through the stratum corneum, lower the impedance between the skin and electrode, reduce motion artifacts, and is convenient for long-term monitoring of bioelectric signals [8,10]. The MAE is a promising alternative to wet electrodes in some specific situations. Various fabrication methods of the MAE are employed to fabricate MA from silicon, metal, or polymer. Photolithography technology with wet etch or dry etch is always used to fabricate MA on a silicon substrate [5,10,13–15]. This method is fit for mass production of MA. However, it has disadvantages: firstly, photolithography or etching needs sophisticated equipment located in a clean room and may produce toxic waste. It is inconvenient, expensive, and eco-unfriendly. Secondly, silicon micro-needles may easily break off and stay in the skin due to its fragile property [5,15]. Thus, metal is adopted to fabricate MA due to its good strength. Nanosecond IR pulsed-laser machining was used to fabricate MA on a pure copper substrate for bio-signals monitoring [3]. Nanosecond pulsed-laser machining has high machining efficiency and flexibility. However, the surface of the micro-needles is usually rough due to the recast layer and debris deposition on it, and the tips are relatively blunt. The biocompatibility of pure copper also needs to be further discussed. Polymer is proposed to fabricate MA due to its good toughness and high yield for the skin insertion of MA. Vacuum-casting technology was employed to fabricate multiple micro-spike electrodes for EEG recording [11]. The master pattern was made by CNC micromachining. This technology can also be applied in mass production while its process is relatively complex. 3D printing technology was also proposed to fabricate 3D MA using biocompatible acrylic-based resin for EEG and ECG recording [6]. This technique allowed for personalized customization and flexibility. However, the resolution of 3D printing was low, the size of MA was in the millimeter scale and the tip diameter was about 100 µm. As the MA was fabricated, a conductive film, such as Ti, Au, Ag, or AgCl, should be subsequently coated on MA by sputtering electroless plating, electrolysis, e-beam evaporation, etc. [5,10,13,16,17]. In this paper, we introduce a thermal drawing method to fabricate 3D MA from 2D planar thermosetting polymer film. A simple thermal drawing setup was designed and fabricated. The polymer is heated to a glassy state, drawn into 3D MA by pillar arrays, and finally solidified by thermal curing. Thermal drawing technology is very simple, efficient, and low cost. The technology is also fit for various polymers and mass fabrication. Subsequently, MA is coated with a conductive film of Ti/Au by magnetron sputtering to fabricate the MAE. The biocompatibility of the MAE will be discussed. EII of the MAE and Ag/AgCl electrodes will be measured under compression force. Finite element methods will be proposed to simulate the insertion and pull process of the MAE. The performance of monitoring ECG, EMG, and EEG bio-signals will be evaluated in comparison with commercial Ag/AgCl electrodes. 2. Experimental 2.1. MAE Fabrication 2.1.1. Experimental Setup A simple and effective experimental setup was designed and fabricated for thermal drawing of MA, as shown in Figure 1. It consists of a temperature adjustment module, z-axis positioner, and stainless steel pillar array. In the temperature adjustment module, heating rods are employed to heat copper blocks at a given temperature, and constant temperature water bath pumps low-temperature water to rapidly cool the bottom copper block. The z-axis positioner could manually slip along the z direction with the pillar array to draw glassy-state polymer into a 3D MA.

Sensors 2016, 16, 908 Sensors 2016, 16, 908

3 of 13 3 of 13

Figure 1. 1. Thermal Thermaldrawing drawing setup setup for for MA. MA. Figure

2.1.2. Fabrication Fabrication Process Process 2.1.2.

Materials Materials Preparation Preparation Thermosetting Thermosetting polymer, polymer, 75/25 75/25PLGA PLGA(Mw (Mw==200 200kDa) kDa)was waspurchased purchasedfrom fromShenzhen Shenzhen Polymtek Polymtek Biomaterial Co., Ltd. (ShenZhen, China) PLGA has been widely used in tissue scaffolds and Biomaterial Co., Ltd. (ShenZhen, China) PLGA has been widely used in tissue scaffolds and is is safe safe for round PLGA film of 0.6 mmmm thickness andand 10 mm diameter was cut. for skin skinpenetration penetration[18]. [18].A A round PLGA film of 0.6 thickness 10 mm diameter wasThe cut. material of the pillar array was 316L stainless steel, the height was 3 mm, and its diameter was The material of the pillar array was 316L stainless steel, the height was 3 mm, and its diameter was 350 × 66 with 350 μm. µm. The The pillars pillars were were welded welded on on the the upper upper copper copper block block in in an an array array of of 66 ˆ with aa separation separation of of 11 mm. mm. Thermal Drawing of MA Thermal Drawing of MA The thermal drawing process of MA is shown in Figure 2a–d. Firstly, the round PLGA film was The thermal drawing process of MA is shown in Figure 2a–d. Firstly, the round PLGA film fixed to the bottom copper block and heated at 140 °C. The PLGA film was heated into a glassy state, was fixed to the bottom copper block and heated at 140 ˝ C. The PLGA film was heated into a glassy as shown in Figure 2a. The upper copper block was heated at 160 °C to preheat the pillar array. state, as shown in Figure 2a. The upper copper block was heated at 160 ˝ C to preheat the pillar Secondly, the pillar array was moved down towards the glassy-state PLGA film until the pillar array array. Secondly, the pillar array was moved down towards the glassy-state PLGA film until the pillar touched the PLGA film, as shown in Figure 2b. Thirdly, the pillar array was moved upward at a array touched the PLGA film, as shown in Figure 2b. Thirdly, the pillar array was moved upward at speed of 0.25 mm/s and then necks were formed between the pillar array and PLGA film due to the a speed of 0.25 mm/s and then necks were formed between the pillar array and PLGA film due to surface tension and became narrower. Fourthly, the bottom copper block was rapidly cooled to room the surface tension and became narrower. Fourthly, the bottom copper block was rapidly cooled to temperature (about 25 °C), as shown in Figure 2c. Finally, the pillar arrays were move upward room temperature (about 25 ˝ C), as shown in Figure 2c. Finally, the pillar arrays were move upward continuously and the necks were broken due to the high temperature of the pillar array. MA tips continuously and the necks were broken due to the high temperature of the pillar array. MA tips were were formed, in Figure 2d. Sensors 2016, 16,as 908shown 4 of 13 formed, as shown in Figure 2d. MA Coating 10 nm Ti film and 100 nm Au film were uniformly coated on the whole MA surface in sequence by a magnetron sputtering machine (MSP-3300, Beijing Jinshengweina Technology Co., Ltd, Beijing, China) as shown in Figure 2e. The coating Ti film was used to increase the bonding strength between the PLGA and Au film [6]. The coating Au film could insure the conductivity of the MAE [6,14]. The coated micro-needle array was observed by scanning electron microscopy (SEM, JSM-6380LA, JEOL, Tokyo, Japan). MAE Assembling An electrode button was glued to the backside of the MA by conductive silver glue and a medical Figure 2.2.Fabrication MAE. thermal drawing process of MA, MA,as (e)Sputtering Sputtering coating adhesive dressing was process bonded standard-shape snapprocess connector, shown incoating Figure 2f. Figure Fabrication processof ofwith MAE.a(a)–(d): (a)–(d): thermal drawing of (e) Ti/Au on the of MAE. The medical adhesive dressing could insure MAE sticks on, and penetrates into, the skin. The standard Ti/Aufilm film onthe the MA, MA, and and (f) (f) the assembly assembly of MAE. snap connector could guarantee the connection between the MAE and common bio-signals 2.2. Insertion, EII Test, and Numerical Simulation recording devices. 2.2.1. EII Test during the Insertion Process The MAE can penetrate through the stratum corneum layer and eliminate the impendence of the dead skin to collect EII. We designed a setup to test the EII of the MAE during the insertion process, as shown in Figure 3. The left inner forearm was chosen as the measurement object due to

Sensors 2016, 16, 908

4 of 13

MA Coating Sensors 2016, 16, 908

4 of 13

10 nm Ti film and 100 nm Au film were uniformly coated on the whole MA surface in sequence by a magnetron sputtering machine (MSP-3300, Beijing Jinshengweina Technology Co., Ltd, Beijing, China) as shown in Figure 2e. The coating Ti film was used to increase the bonding strength between the PLGA and Au film [6]. The coating Au film could insure the conductivity of the MAE [6,14]. The coated micro-needle array was observed by scanning electron microscopy (SEM, JSM-6380LA, JEOL, Tokyo, Japan). MAE Assembling An electrode button was glued to the backside of the MA by conductive silver glue and a medical adhesive dressing was bonded with a standard-shape snap connector, as shown in Figure 2f. The medical adhesive dressing could insure MAE sticks on, and penetrates into, the skin. Figure The standard snapprocess connector could guarantee connection the MAE and common 2. Fabrication of MAE. (a)–(d): thermalthe drawing processbetween of MA, (e) Sputtering coating bio-signals recording devices. Ti/Au film on the MA, and (f) the assembly of MAE. 2.2. 2.2. Insertion, Insertion, EII EII Test, Test, and and Numerical Numerical Simulation Simulation 2.2.1. EII Test during the Insertion Process 2.2.1. EII Test during the Insertion Process The MAE can penetrate through the stratum corneum layer and eliminate the impendence of The MAE can penetrate through the stratum corneum layer and eliminate the impendence of the dead skin to collect EII. We designed a setup to test the EII of the MAE during the insertion the dead skin to collect EII. We designed a setup to test the EII of the MAE during the insertion process, as shown in Figure 3. The left inner forearm was chosen as the measurement object due to process, as shown in Figure 3. The left inner forearm was chosen as the measurement object due to having less hair, a thinner stratum corneum, as well as being a convenient and accurate position for having less hair, a thinner stratum corneum, as well as being a convenient and accurate position for electrodes [3,19]. A two-electrode measurement method was proposed to record EII [3]. The location of electrodes [3,19]. A two-electrode measurement method was proposed to record EII [3]. The location the MAE and Ag/AgCl electrode is shown in Figure 3. The electrodes were connected to the precision of the MAE and Ag/AgCl electrode is shown in Figure 3. The electrodes were connected to the impedance analyzer (Agilent E4980A LCR Meter, Palo Alto, CA, USA) with coaxial wires, to avoid precision impedance analyzer (Agilent E4980A LCR Meter, Palo Alto, CA, USA) with coaxial wires, noise. A linear motor (E-861, PI, Karlsruhe, Baden-Württemberg, German) could load the MAE on to avoid noise. A linear motor (E-861, PI, Karlsruhe, Baden-Württemberg, German) could load the the forearm and the force sensor (Nano 17 Titanium, ATI Industrial Automation, Detroit, MI, USA) MAE on the forearm and the force sensor (Nano 17 Titanium, ATI Industrial Automation, Detroit, captured the insertion force. The EII signal, insertion force, and displacement could be simultaneously MI, USA) captured the insertion force. The EII signal, insertion force, and displacement could be recorded by self-developed software during the insertion process. simultaneously recorded by self-developed software during the insertion process. In order to better understand the effect of insertion force on the recording of EII, the EII of the In order to better understand the effect of insertion force on the recording of EII, the EII of the MAE and insertion force were recorded during the insertion process. The linear motor speed was MAE and insertion force were recorded during the insertion process. The linear motor speed was 0.5 mm/s, the injection voltage of the LCR meter was set at 1 V, and its frequency was set at 50 Hz. 0.5 mm/s, the injection voltage of the LCR meter was set at 1 V, and its frequency was set at 50 Hz. We also recorded EII with the frequency from 20 Hz to 10 kHz when the insertion force was held at We also recorded EII with the frequency from 20 Hz to 10 kHz when the insertion force was held at a constant value (about 1 N, 2 N, or 3 N). As the measured impedance is beyond 2 MΩ, we set the test a constant value (about 1 N, 2 N, or 3 N). As the measured impedance is beyond 2 MΩ, we set the result as 2 MΩ. test result as 2 MΩ.

Figure 3. Setup for EII recording during the insertion insertion process. process.

2.2.2. Numerical Simulation of Insertion and Pull Process The mechanism of the MAE insertion and pull process to human skin has not yet been understood clearly. We proposed a nonlinear FEM model to simulate this process by ABAQUS Explicit as shown in Figure. 4. For simplicity, we simulated it with a 2-D plane-strain model under quasi-static conditions. The MAE was inserted into the skin at a vertical direction and drawn from the skin at a speed of 50 μm/s. The insertion depth was 150 μm.

Sensors 2016, 16, 908

5 of 13

2.2.2. Numerical Simulation of Insertion and Pull Process The mechanism of the MAE insertion and pull process to human skin has not yet been understood clearly. We proposed a nonlinear FEM model to simulate this process by ABAQUS Explicit as shown in Figure 4. For simplicity, we simulated it with a 2-D plane-strain model under quasi-static conditions. The MAE was inserted into the skin at a vertical direction and drawn from the skin at a speed of 50 µm/s. The insertion depth was 150 µm. Human skin consists of four layers: stratum corneum, epidermis, dermis, and hypodermis. We assumed the skin of each layer as the isotropous and incompressible material to simplify the SensorsThe 2016, 16, 908 of 13 simulation. Neo-Hooken constitutive model of skin was employed [20,21] and the 5mechanical properties are listed in Table 1 [22–24]. The hypodermis layer was ignored in the model since the MAE Human skin consists of four layers: stratum corneum, epidermis, dermis, and hypodermis. We could not reach the andskin its effect was minimal. The material of skinmaterial was defined by thetheVUMAT assumed of each layer as the isotropous and model incompressible to simplify subroutine in ABAQUS. The failure criterionmodel handled by was the distortion energy theory was introduced simulation. The Neo-Hooken constitutive of skin employed [20,21] and the mechanical properties are listed in Table 1 [22–24]. The hypodermis layer was ignored themicro-needle model since thetips was in the skin material model. When the effective stress of an element nearinthe could not failure reach and its effectitwas minimal. The material of skin was by the beyondMAE the specified criterion, would be identified andmodel eliminated fromdefined the mesh. The skin VUMAT subroutine in ABAQUS. The failure criterion handled by the distortion energy theory was was defined as a deformable body and meshed with 12,000 CPE4R elements, each micro-needle was introduced in the skin material model. When the effective stress of an element near the micro-needle definedtips as awas discrete rigid body and meshed with 469 CPS4R elements, and the base of the MAE was beyond the specified failure criterion, it would be identified and eliminated from the definedmesh. as anThe analytical body. skin wasrigid defined as a deformable body and meshed with 12,000 CPE4R elements, each micro-needle was defined as a discrete rigid body and meshed with 469 CPS4R elements, and the Table properties base of the MAE was defined as 1. anMechanical analytical rigid body. of human skin. Stratum properties Corneum of human Epidermis Table 1. Mechanical skin. C10 (MPa) Failure stress (MPa) C10 (MPa) Density (kg/m3 ) Failure stress (MPa) Thickness (mm) Density (kg/m3) Thickness (mm)

10Corneum Stratum 25 10 1300 25 0.07 1300 0.07

0.11 Epidermis – 0.11 1200 -0.05 1200 0.05

Dermis

Dermis0.16 7.3 0.16 1200 7.3 0.1 1200 0.1

Surface-to-surface contact was defined between micro-needles and skin using the kinematic Surface-to-surface contact was defined micro-needles andwith skin using the kinematic contact algorithm. Contact between the MAEbetween and skin was modeled the friction coefficient of contact algorithm. Contact between the MAE and skin was modeled with the friction coefficient of 0.42 [25]. The rotation of the X, and X and Y translation were constrained in the nodes of the MAE. 0.42 [25]. The rotation of the X, and X and Y translation were constrained in the nodes of the MAE. The displacement boundary conditions included nodes along the bottom, left, and right The displacement boundary conditionson onthe the skin skin included nodes along the bottom, left, and right edges that were pinned, whereas the top edge was traction-free. The explicit method was introduced edges that were pinned, whereas the top edge was traction-free. The explicit method was introduced to calculate the process. to calculate the process.

Figure 4. Finite element model of the MAE insertion and pull process. Figure 4. Finite element model of the MAE insertion and pull process.

2.3. Bio-Signals Recording

2.3. Bio-Signals Recording

In order to better understand the sensing performance of MAE, bio-signals, including EMG,

In order to better understand thethesensing of conventional MAE, bio-signals, ECG, and EEG, were recorded by MAE inperformance comparison with Ag/AgClincluding electrodes EMG, (JK- ECG, 1(A~H) Shanghai Junkang Supplies LTD., CO, Shanghai, China). Measurements and EEG, weretype, recorded by the MAEMedical in comparison with conventional Ag/AgCl electrodeswere (JK-1(A~H) firstly performed with the MAE and thenLTD., repeated Ag/AgClChina). electrodes under the samewere test firstly type, Shanghai Junkang Medical Supplies CO,with Shanghai, Measurements conditions. The electrodes were located at the same position and followed the same procedures in performed with the MAE and then repeated with Ag/AgCl electrodes under the same test conditions. both cases. The experiments were operated on three healthy volunteers from 23 to 26 years old at The electrodes were located at the at same room temperature and repeated least position five times. and followed the same procedures in both cases. The experiments were operated on three healthy volunteers from 23 to 26 years old at room temperature 2.3.1. EMG Test five times. and repeated at least The biceps brachii muscle of the right arm was employed as the EMG measuring object, as shown in Figure 5. The differential method was introduced to record EMG signals. The upper arm was placed on the plate and the right hand held the stick. The stick was assembled on one end of

Sensors 2016, 16, 908

6 of 13

2.3.1. EMG Test The biceps brachii muscle of the right arm was employed as the EMG measuring object, as shown in Figure 5. The differential method was introduced to record EMG signals. The upper arm was placed on the plate right hand held the stick. The stick was assembled on one end of plate. Sensorsand 2016, the 16, 908 6 ofThe 13 angle between upper arm and forearm was 90˝ . If the upper arm tried to rotate the plate, the torque sensor plate. The angle between upper arm and forearm was 90°. If the upper arm tried to rotate the plate, (AKC-205, 701st Research Institute of China Aerospace Science and Technology Corporation, Beijing, the torque sensor (AKC-205, 701st Research Institute of China Aerospace Science and Technology China) are the otherBeijing, end ofChina) the plate would value. Twothe recording electrodes Corporation, are the otherrecord end of the the torque plate would record torque value. Two were stuck onrecording the biceps brachii with a distance of 2 cm and a ground electrode was placed on the elbow. electrodes were stuck on the biceps brachii with a distance of 2 cm and a ground electrode These electrodes were connected to a tele-EMG system (MyoSystem2400T, Noraxon, Scottsdale, AZ, was placed on the elbow. These electrodes were connected to a tele-EMG system (MyoSystem2400T, Noraxon, Scottsdale, AZ, USA). EMG signals were collected by data acquisition card (DAQ-6341, USA). EMG signals were collected by data acquisition card (DAQ-6341, National Instruments, Austin, National Instruments, TX, USA) and analyzed by (LabVIEW a customized LabVIEW program TX, USA) and analyzed by aAustin, customized LabVIEW program 2012, National Instruments (LabVIEW 2012, National Instruments Corporation, Austin, TX, USA). The skin was firstly cleaned Corporation, Austin, TX, USA). The skin was firstly cleaned with medical alcohol. The volunteer held with medical alcohol. The volunteer held the stick and increased the torque from 0 Nm to 15 Nm. As the sticktheand increased torque 0 Nm toheld 15 for Nm. As5the torque torque reached the about 15 Nm,from the volunteer about s. Then, the reached volunteer about released15 theNm, the volunteer held for about 5 s. Then, the volunteer released the force. The torque and EMG signals force. The torque and EMG signals were recorded simultaneously. The process was be repeated were several times. recorded simultaneously. The process was be repeated several times.

Figure 5. Schematicdiagram diagram of measurement. Figure 5. Schematic ofEMG EMG measurement.

2.3.2. ECG Test

2.3.2. ECG Test

The MAE sensing performance of ECG signals was operated in static and dynamic state by

Thestandard MAE sensing performance ofwere ECGconnected signals was in staticfrom andMultipurpose dynamic state by II-lead method. Electrodes to an operated ECG100C module Polygraph BIOPAC, Goleta,were CA, USA). Measuring electrodes were module stuck on the rightMultipurpose wrists standard II-lead (MP150, method. Electrodes connected to an ECG100C from and (MP150, left ankle, BIOPAC, and grounded electrode on the right ankle.electrodes The volunteer laystuck in a bed the wrists Polygraph Goleta, CA,was USA). Measuring were onduring the right static state test and walked on a treadmill at a uniform velocity of 3 km/h during the dynamic state and left ankle, and grounded electrode was on the right ankle. The volunteer lay in a bed during the test. The test lasted about half an hour. static state test and walked on a treadmill at a uniform velocity of 3 km/h during the dynamic state test. The2.3.3. testEEG lasted Testabout half an hour. EEG measures voltage fluctuations resulting from ionic current within the neurons of the brain. 2.3.3. EEG Test

A unipolar connection method was used in the EEG test. Three electrodes were connected to the

EEG100C module of Multipurpose Polygraph, twofrom measuring electrodeswithin and one ground electrode. EEG measures voltage fluctuations resulting ionic current the neurons of the brain. One measuring electrode was placed on the standard position (Fp1) of the 10–20 system. The ground A unipolar connection method was used in the EEG test. Three electrodes were connected to the electrode and the other measuring electrode were located on the left earlobe. The volunteers were EEG100C module of Multipurpose Polygraph, two measuring electrodes and one ground electrode. asked to blink for 30 s and rest for 30 s during the blinking test. Volunteers then were required to One measuring electrode was placed on the standard position the 10–20 system. The alternate closed and opened eyes with an interval of two seconds(Fp1) underof the tester’s guide during the ground electrode and theand other measuring electrode were onabout the left earlobe. The volunteers were eyes closed open transition test. The whole test located process last 5 min. asked to blink for 30 s and rest for 30 s during the blinking test. Volunteers then were required to alternate closed and opened eyes with an interval of two seconds under the tester’s guide during the eyes closed and open transition test. The whole test process last about 5 min.

Sensors 2016, 2016, 16, 16, 908 908 Sensors

of 13 13 77 of

3. Results and Discussion 3. Results and Discussion 3.1. Characterization of the MAE 3.1. Characterization of the MAE The MAE fabricated by thermal drawing and coated with Ti/Au film is shown in Figure 6a. The MAE fabricated by thermal drawing and PLGA, coatedTi, with is shown in Figure 6a. MAEThe mainly consists of PLGA MA and its Ti/Au film. andTi/Au Au arefilm all biocompatible which can The MAE mainly consists of PLGA MA and its Ti/Au film. PLGA, Ti, and Au are all biocompatible guarantee the compatibility of the MAE in human skin [18,26–29]. The MA consists of 6 × 6 micro-needles which guarantee the compatibility the with MAEthat in human skin [18,26–29]. Thearray. MA consists of with ancan interval of 1 mm, which matchesofwell of the stainless steel pillar The height 6ofˆMA 6 micro-needles with an interval of in 1 mm, which matches wellheight with that of ±the steel distributes uniformly as shown Figure 6b. The average is 500 10 stainless μm, which is pillar array. The height of MA distributes uniformly as shown in Figure 6b. The average height is appropriate for painless recording [10]. The MA surface was very smooth, which may decrease the 500 ˘ 10friction µm, which appropriate recording [10]. The MA surface was veryofsmooth, which sliding forceisbetween skinfor andpainless MA during the penetration. One micro-needle MA is shown may decrease the sliding friction force between skin and MA during the penetration. One micro-needle in Figure 6c. The shape of the micro-needle looks like a cone. The base diameter of micro-needles was of MA500 is shown in and Figure Theof shape of the micro-needle cone. The base diameter of about ± 10 μm its 6c. radius tip curvature was about looks 40 ± 2 like μm.aThe base diameter was kept micro-needles was about 500 ˘ 10 µmand andavoid its radius of tip curvature wasinsertion. about 40 ˘ 2 µm. The base large to strengthen the micro-needle buckling during the skin diameter was kept large to strengthen the micro-needle and avoid buckling during the skin insertion.

Figure 6. (a) Photo of the MAE; (b) SEM image of MA; and (c) SEM image of the micro-needles. Figure 6. (a) Photo of the MAE; (b) SEM image of MA; and (c) SEM image of the micro-needles.

3.2. EII Test and Insertion Process 3.2. EII Test and Insertion Process 3.2.1. EII Recording during MAE Compression 3.2.1. EII Recording during MAE Compression The EII of MAE during the insertion process at 50 Hz is shown in Figure 7a. The EII value is The EII ofatMAE during the insertion process at 50 Hz is shown in Figure EII between value is extremely high the beginning and the compression force is null due to the absence7a. of The contact extremely high at Subsequently, the beginningmicro-needle and the compression force is null to the absence contact the MAE and skin. tips come in contact withdue forearm skin and theofinsertion between the MAE and skin. Subsequently, micro-needle tips come in contact with forearm skin and force gradually increases, and EII maintains its high impedance value due to the high impedance the insertion force gradually increases, and EII maintains its high impedance value due to the high of the stratum corneum layer. As the compression force is about 75 mN, EII suddenly changes the impedance of be thedue stratum layer. As the compression force is about 75 mN, EII suddenly slope. It may to thecorneum penetration of one micro-needle through the stratum corneum and the changes the slope. It may be due to the penetration of one micro-needle through the stratum corneum contact impedance between the MAE and skin decreases rapidly. As the MAE is moved forward, the and the contact betweeninto the the MAE and decreases As the MAE moved micro-needles of impedance the MAE penetrates skin oneskin by one and therapidly. EII decreases with theisinsertion forward, the micro-needles of the MAE penetrates into the skin one by one and the EII decreases depth. When a force larger than 0.55 N is applied on the MAE, the MAE achieves lower EII thanwith the the insertion depth. When a force larger 0.55 N is force applied MAE, the MAE lower Ag/AgCl electrodes (120 KΩ). When thethan compression on on thethe MAE is larger thanachieves 1 N, EII of the EII than the Ag/AgCl KΩ). When the85compression force on the of MAE is larger than 1 MAE reaches a steadyelectrodes state, and(120 its value is about KΩ. The average force an adult applying N, EII of with the MAE steady20 state, andTherefore, its value iswe about KΩ.press The average force an adult the MAE their reaches thumb isa about N [30]. can 85 easily the MAE intoofskin with applying the MAE with their thumb is about 20 N [30]. Therefore, we can easily press the MAE into skin a thumb and stably record the bio-signals. with The a thumb stably record the bio-signals. MAEand impedances of skin-electrode under different compression forces at the driving current The MAE impedances of skin-electrode under different compression forces at thewith driving frequency from 20 Hz to 10 kHz are shown in Figure 7b. The EII of MAE decreases the current frequency from 20 Hz to 10 kHz are shown in Figure 7b. The EII of MAE decreases with the frequency. The impendence of the human body always consist of resistance and capacitive resistance. frequency. The impendence of the human always consist of resistance and capacitive resistance. The capacitive resistance decreases with body the frequency of inject driving current. The EII slightly The capacitive resistance decreases with the frequency of inject driving current. The EII slightly decreases with the insertion force at a given driving current frequency, as shown in Figure 7b. decreases withEII the insertion forceofat a given driving current frequency, as and shown in impedance. Figure 7b. The measured usually consists contact impedance, electrode impedance, tissue The measured EII usually consiststhe of contact impedance, impedance, andthe tissue impedance. The contact impedance between electrode and skin electrode slightly decreased with insertion force The contact impedance between the electrode and skin slightly decreased with the insertion force since both the electrode impedance and tissue impedance maintain a constant value at a given current since both the electrode tissue impedance maintain a constant at a given frequency. Therefore, theimpedance MAE can and record EII or bio-signals at a relatively lowvalue insertion force.current As the frequency. Therefore, the MAE can record EII or bio-signals at a relatively low insertion force. As the the compression force on the MAE is beyond about 5 N, the subjects would feel tingling during compression force on the MAE is beyond about 5 N, the subjects would feel tingling during the

Sensors 2016, 16, 908 Sensors Sensors 2016, 2016, 16, 16, 908 908

of 13 13 888 of of 13

insertion process. Since the MAE can stably record EII aa compression force of 11 N, insertionprocess. process.Since Since MAE stably EII under compression N, the the insertion thethe MAE can can stably recordrecord EII under aunder compression force of 1force N, theofbio-signals bio-signals recording on the subjects is almost painless. The skin deformation under the insertion bio-signals recording on the subjects is almost painless. The skin deformation under the insertion recording on the subjects is almost painless. The skin deformation under the insertion force of 6 N force of shown in Figure 7c. The mark on forearm skin by the will disappear force of 66 N N is is shown Figure 7c.on The mark on the the forearm bywill the MAE MAE will gradually gradually is shown in Figure 7c.inThe mark the forearm skin by theskin MAE gradually disappeardisappear in about in about 8 min and the harm on skin is minimal. about minharm and the harmison skin is minimal. 8inmin and8 the on skin minimal.

(a) EII during the insertion process, Figure Figure 7. 7. Insertion Insertion force force and and EII EII test. test. (a) (a) EII EII during during the the insertion insertion process, process, (b) (b) EII EII under under different different input input frequency, and (c) skin deformation under the impression force of 6N. current frequency, and (c) skin deformation under the impression force of 6N. current frequency, and (c) skin deformation under the impression force of 6N.

3.2.2. 3.2.2. Numerical Numerical Simulation Simulation of of the the Insertion Insertion and and Pull Pull Process Process Figure of the insertion process. As tips 8a–c presentthe theskin skinstress stressdistribution distribution the insertion process. As the micro-needle Figure 8a–c 8a–c present present the skin stress distribution ofof the insertion process. As the the micro-needle micro-needle tips touch the layer of the stress concentration occurs at the tips touch the layer of skin, stress concentration occurs thetip tiparea areaof ofMAE MAE as as shown shown touch the top top top layer of skin, skin, thethe stress concentration occurs atat the tip area of MAE as shown in in Figure 8a. Top layer skin initially deforms concave downward under the advancing tips until 8a. Top layer skin initially deforms concave downward under the advancing tips until a critical Figure 8a. Top layer skin initially deforms concave downward under the advancing tips until aa critical stress to effective stress of elements the stress to leads the skin penetration. Once the Once effective of skin elements near the tipsnear exceeds the criticalleads stress leads to the the skin skin penetration. penetration. Once the thestress effective stress of skin skin elements near the tips tips exceeds specified failure criterion, the nodes separated due specified failure criterion, the nodes are to the deletion ofdeletion “dead” of elements. Thus, the exceeds the the specified failure criterion, theseparated nodes are aredue separated due to to the the deletion of “dead” “dead” elements. elements. Thus, “tear” skin. The skin is on micro-needle tips “tear” tips the skin. The skin stress is always on the micro-needle Thus, the the micro-needle micro-needle tipshuman “tear” the the human human skin. Theconcentration skin stress stress concentration concentration is always always on the the micro-needle tips during the insertion process, which verifies that the skin tissue is cut away by tips during thetips insertion whichprocess, verifieswhich that the skin tissue is cut byisthe as micro-needle duringprocess, the insertion verifies that the skinaway tissue cutMAE awaytips, by the the MAE tips, as shown in Figure 8a–c. The stress concentration at the micro-needle surface may lead shown in Figure 8a–c. The stress concentration at the micro-needle surface may lead to the increase MAE tips, as shown in Figure 8a–c. The stress concentration at the micro-needle surface may lead to to the increase of the force between friction force the of friction between micro-needles and skin. and The friction increases with thewith contact thethe increase of force the friction friction force between micro-needles micro-needles and skin. skin. The Theforce friction force increases increases with the contact area micro-needles and as MAE skin more fits well area between micro-needles and skin asskin the MAE in the in skin It fits It well contact area between between micro-needles and skin as the theinserts MAE inserts inserts in the themore skin deeply. more deeply. deeply. It fitswith well with the plot of insertion force vs. displacement is shown in Figure 7a. The stress interaction the plot of insertion force vs. displacement is shown in Figure 7a. The stress interaction of adjacent with the plot of insertion force vs. displacement is shown in Figure 7a. The stress interaction of of adjacent can ignored during the process Figure 8a–c, so micro-needles can be ignored insertion process as shownas inshown Figurein 8a–c, so the distance adjacent micro-needles micro-needles can be beduring ignoredthe during the insertion insertion process as shown in Figure 8a–c, so the the distance of micro-needles (1 mm) for MAE 8c–e the of adjacent micro-needles (1 mm) is for MAE penetration. Figure Figure 8c–e present the stress distance of adjacent adjacent micro-needles (1suitable mm) is is suitable suitable for MAE penetration. penetration. Figure 8c–e present present the stress distribution of the pull process. As the MAE is drawn back, the compression state of the skin distribution of the pull process. As the MAE is drawn back, the compression state of the skin near the stress distribution of the pull process. As the MAE is drawn back, the compression state of the skin near into aa tensile The profile of deforms from to micro-needles turns intoturns a tensile Thestate. profile of skin deforms from concave to concave convex, and finally near the the micro-needles micro-needles turns intostate. tensile state. The profile of skin skin deforms from concave to convex, convex, and finally becomes flat. The taper holes gradually shrink due to the elasticity of skin during the becomes flat. The taper holes gradually shrink due to the elasticity of skin during the pull process. and finally becomes flat. The taper holes gradually shrink due to the elasticity of skin during the pull pull process. The fits the The shape of shape insertion holes fitsholes with the skin in Figurein 7c.Figure process. The shape of of insertion insertion holes fits with withdeformation the skin skin deformation deformation in Figure 7c. 7c.

Figure and pull process. Figure 8. 8. Stress Stress distribution distribution on on human human skin skin during during the the insertion insertion and and pull pull process. process. Figure 8. Stress distribution on human skin during the insertion

Sensors 2016, 16, 908

9 of 13

Sensors 2016, 16, 908

9 of 13

3.3. Bio-Signals Measurement 3.3. Bio-Signals Measurement 3.3.1. EMG Measurement 3.3.1. EMG Measurement EMG signal has a variety of biomedical applications, such as a diagnostics tool for identifying EMG signal has a avariety of tool biomedical applications, such aascontrol a diagnostics toolprosthetic for identifying neuromuscular diseases, research for studying kinesiology, signal for devices, a research tool for studying kinesiology, a control for prosthetic andneuromuscular so on. Surface diseases, EMG signals of the biceps brachii muscle recorded by the MAEsignal is shown in Figure 9a devices, on. Surface signals the biceps byelectrical the MAE potential is shown of as the rightand armsoapplied forceEMG to rotate the of stick. As the brachii torque muscle is aboutrecorded 0 Nm, the in Figure 9a as the right arm applied force to rotate the stick. As the torque is about 0 Nm, the electrical biceps brachii muscle cells is in a low amplitude. When the torque increases from 0 Nm to 15 Nm, the potential biceps brachii or muscle cells is in aactivated, low amplitude. When the torque increases from 0rapidly Nm muscle cellsofare electrically neurologically and the electrical potential increases to 15 Nm, the muscle cells are electrically or neurologically activated, and the electrical potential to a high amplitude. The biceps brachii muscle is contracted from the relaxation state. When the increases rapidly to a high amplitude. The biceps brachii muscle is contracted from the relaxation volunteer holds the stick with the torque of 15 Nm for 5 s, the EMG signal maintains a high value state. When the volunteer holds the stick with the torque of 15 Nm for 5 s, the EMG signal maintains due to the sustained contraction of the muscle. The biceps brachii muscle is released and the EMG a high value due to the sustained contraction of the muscle. The biceps brachii muscle is released and signal is at a low amplitude again as the rotation force is unloaded. The EMG recording process the EMG signal is at a low amplitude again as the rotation force is unloaded. The EMG recording by the MAE suggests that the MAE can trace the change of EMG signals. The EMG signal of the process by the MAE suggests that the MAE can trace the change of EMG signals. The EMG signal of biceps brachii muscle recorded by the conventional Ag/AgCl electrode also shows a similar profile in the biceps brachii muscle recorded by the conventional Ag/AgCl electrode also shows a similar comparison with the MAE presented Figure 9.inItFigure demonstrates that the MAE canMAE clearly and profile in comparison with the MAE in presented 9. It demonstrates that the cansense clearly record surface signalEMG with signal good fidelity. The contact surface ofsurface the conventional Ag/AgCl sensethe and recordEMG the surface with good fidelity. The contact of the conventional 2 and the MAE is only about 78.5 mm2 . The MAE is more suitable for use electrode is about 200 mm 2 2 Ag/AgCl electrode is about 200 mm and the MAE is only about 78.5 mm . The MAE is more suitable on for theuse small muscles. Furthermore, micro-needles distribution, number, andand sizesize of the MAE can on the small muscles. Furthermore, micro-needles distribution, number, of the MAE be can designed and customized by thermal drawing method according to thetoshape and size muscle be designed and customized by thermal drawing method according the shape andofsize of monitored. Thus, the Thus, MAE may be more for both improvement of EMG signal selectivity muscle monitored. the MAE mayconvenient be more convenient for both improvement of EMG signal and elimination the crosstalk. Thiscrosstalk. is a critical phenomenon the case of several muscles present selectivity andofelimination of the This is a critical in phenomenon in the case of several present in a small space or when volunteers have small anthropometric dimensions. in amuscles small space or when the volunteers havethe small anthropometric dimensions. Therefore, the MAE Therefore, MAEin is EMG also a recording good choice EMG recording forapplications potential medical due to is also a goodthe choice forinpotential medical due toapplications its good sensitivity its good sensitivity and fidelity. and fidelity.

Figure 9. 9. EMG (b) Ag/AgCl Ag/AgClelectrodes. electrodes. Figure EMGsignals signalsrecorded recordedby: by: (a) (a) MAE; MAE; and (b)

3.3.2. ECG Measurement 3.3.2. ECG Measurement ECG signal is always used test frequency rate heartbeats,the thesize sizeand andposition positionofofthe AnAn ECG signal is always used to to test thethe frequency rate ofof heartbeats, the heart chambers, the presence of any damage to the heart's muscle cells, the effects of cardiac heart chambers, the presence of any damage to the heart's muscle cells, the effects of cardiac drugs, drugs, and the function of implanted pacemakers [31]. Figure 10a presents 4-s ECG recording results and the function of implanted pacemakers [31]. Figure 10a presents 4-s ECG recording results using using the MAE and Ag/AgCl electrodes from the static state test. The recording performance of the the MAE and Ag/AgCl electrodes from the static state test. The recording performance of the MAE MAE is comparable with Ag/AgCl electrodes. The features of the ECG signal, such as the QRS is comparable with Ag/AgCl electrodes. The features of the ECG signal, such as the QRS complex, complex, T and P waves, are all distinguishable as reported in [2,7,32]. The heart rate in this subject T and P waves, are all distinguishable as reported in [2,7,32]. The heart rate in this subject is about is about 65 beats per minute. Therefore, the MAE can record the characteristic ECG peaks effectively 65 beats per minute. Therefore, the MAE can record the characteristic ECG peaks effectively in the in the static state. Figure 10b presents 3-s ECG recording results using the MAE and Ag/AgCl electrodes static state. Figure 10b presents 3-s ECG recording results using the MAE and Ag/AgCl electrodes from the dynamic state test. The ECG signals are seriously affected by motion artifacts [7,33]. As the from the dynamic state test. The ECG signals are treadmill, seriously the affected by motion artifacts As the subject swung the arms during walking on the skin shifted in respect of [7,33]. underlying

Sensors 2016, 16, 908

10 of 13

subject swung the arms during walking on the treadmill, the skin shifted in respect of underlying Sensors 2016, 16, 908 10 of 13 tissues, modifying the electrodes relative positions and generating the signal drifts and motion artifacts. The signals measured by both the Ag/AgCl and the MAE seriously disturbed tissues, modifying the electrodes relativeelectrodes positions and generating theare signal drifts and motion by noise signals measured by both thewith Ag/AgCl and the MAE due to theartifacts. motionThe artifacts in comparison the electrodes signals recorded inare theseriously static disturbed state. The typical by noise due to the motion artifacts in comparison with the signals recorded in the static state. Sensors 2016, 16, 908 10 The of 13 R waves and T waves of the MAE results are recognizable and the heart frequency rate can be calculated. typical R waves and T waves of the MAE results are recognizable and the heart frequency rate can ECG signal collected by has more QRS complex, P drifts waves and T waves than tissues, modifying theMAE electrodes relative positions anddistinguishable generating theQRS signal motion be calculated. ECGthe signal collected by the distinguishable MAE has more complex, Pand waves and artifacts. signals both the Ag/AgCl electrodes andcollected the MAE areAg/AgCl seriously disturbed shift from signal by Ag/AgCl electrodes. The by ECG signals collected by Ag/AgCl electrodes seriously T waves The than signalmeasured by Ag/AgCl electrodes. The ECG signals by electrodes by noise the artifacts in infer comparison signals the static state. The seriously shifttofrom the We thatinto thewith MAE can insert into theinskin maintain a the baseline. We due infer thatmotion thebaseline. MAE can insert thethe skin andrecorded maintain a and relatively more stable typical R waves and Telectrode-skin waves of the contact MAE results are recognizable and the heart frequency rate can relatively more stable interface. electrode-skin contact interface. be calculated. ECG signal collected by the MAE has more distinguishable QRS complex, P waves and T waves than signal by Ag/AgCl electrodes. The ECG signals collected by Ag/AgCl electrodes seriously shift from the baseline. We infer that the MAE can insert into the skin and maintain a relatively more stable electrode-skin contact interface.

Figure 10.signals ECG signals recorded: (a)in in the the static andand (b) in(b) thein dynamic state. Figure 10. ECG recorded: (a) staticstate, state, the dynamic state.

EEG signals can be used to diagnose epilepsy, sleep disorders, coma, encephalopathies, brain

EEG death, signals beThe used epilepsy, disorders, coma, encephalopathies, brain andcan so on. EEGtois diagnose more difficult to capturesleep since its amplitude is in the order of μV while the ECG is in the range of mV. EEG signals recorded on Fp1 by MAE and Ag/AgCl electrodes are Figure 10. ECG signals recorded: (a) in the static state, and (b) in the dynamic state. death, and so on. The EEG is more difficult to capture since its amplitude is in the order of µV while the shown in Figure 11. Signal profiles captured by on the Fp1 MAEby areMAE very similar with that ofelectrodes the Ag/AgClare shown ECG is in the range of mV. EEG signals recorded and Ag/AgCl EEG signals be used to diagnose epilepsy, sleep as disorders, encephalopathies, brain electrodes during can the eyes closed and eyes open transition shown incoma, Figure 11a. Both beta rhythm in Figure 11. Signal profiles by the MAE are very similar withisthat oforder the Ag/AgCl electrodes death, and so on. Thecaptured EEG is more difficult to capture its amplitude in the of μVinwhile (eyes open condition) and alpha rhythm (eyes closed since condition) are recognizable, as found [10]. during theWhen eyes closed and open transition as shown inthe Figure Both rhythm the ECG issubject in the range of mV. signals recorded on Fp1 by blink MAE11a. and Ag/AgCl electrodes are(eyes open the waseyes asked to EEG blink the eyes successively, signal comesbeta in on the EEG shown in Figure 11. Signal profiles captured by the MAE veryobserved similar with that of theMAE Ag/AgCl and rhythm fluctuates regularly. The condition) same fluctuations are byasboth the and When the condition)recording and alpha (eyes closed arearerecognizable, found in [10]. electrodeselectrodes, during theaseyes closed and eyes shown in Figure rhythm Ag/AgCl shown in Figure 11b.open Bothtransition the MAE as Ag/AgCl electrodes can monitor EEG subject was asked to blink the eyes successively, the blinkand signal comes in11a. onBoth thebeta EEG recording and (eyes open and alpha capture rhythm bio-signals (eyes closedwithout condition) recognizable, in [10]. signals, but condition) MAE could directly skinare preparation (suchasasfound hair cutting fluctuates and regularly. The was same fluctuations both thesignal MAE and When the subject asked to blink theare eyesobserved successively, blink comes inAg/AgCl on the EEGelectrodes, skin abrasion) and application of electrolytic gel. It mayby bethe more comfortable for subjects without as shown skin in Figure 11b. Both the MAE and Ag/AgCl electrodes can monitor EEG signals, recording and fluctuates regularly. The same fluctuations are observed by both the MAE and but MAE abrasion, skin allergy, and the bother of the gel drying in long term monitoring [2]. Ag/AgCl electrodes, as shown in Figure 11b. Both the MAE and Ag/AgCl electrodes can monitor EEG could directly capture bio-signals without skin preparation (such as hair cutting and skin abrasion) signals, but MAE could directly capture bio-signals without skin preparation (such as hair cutting and application of electrolytic gel. It may be more comfortable for subjects without skin abrasion, skin and skin abrasion) and application of electrolytic gel. It may be more comfortable for subjects without allergy, and the bother of the geland drying in long term monitoring [2]. monitoring [2]. skin abrasion, skin allergy, the bother of the gel drying in long term

Figure 11. EEG signals recorded by Ag/AgCl electrodes and the MAE: (a) closed and opened eyes transition, and (b) blinking the eyes.

Figure 11. EEG signals recorded by Ag/AgCl electrodes and the MAE: (a) closed and opened eyes

Figure 11. EEG signals recorded by Ag/AgCl electrodes and the MAE: (a) closed and opened eyes transition, and (b) blinking the eyes. transition, and (b) blinking the eyes.

Sensors 2016, 16, 908

11 of 13

4. Conclusions Micro-needle array electrode was developed for bio-signals monitoring in this paper. The insertion process of the MAE in the skin was simulated, and bio-signals recording performance, including EII, EMG, ECG, and EEG, were investigated by experiments. We obtained the main conclusions: (1)

(2)

The MAE, biocompatible for humans, can be fabricated by a thermal drawing method and magnetron sputtering in mass production. The micro-needle tips can “tear” and insert the human skin due to the stress concentration near the tips. When the compression force on the MAE is larger than 1 N, EII of the MAE is about 85 KΩ which is lower than that measured by Ag/AgCl electrodes (120 KΩ). EII slightly decreases with the insertion force at a given driving current frequency. The MAE can record EII at a relatively low insertion force. The MAE can clearly sense and record the surface EMG signals as well as Ag/AgCl electrodes. The geometry of the MAE can be customized by a thermal drawing method based on a monitored muscle to improve the selectivity of EMG and eliminate the crosstalk. The ECG features measured by both the MAE and Ag/AgCl electrodes are all distinguishable in the static state. ECG signals in the dynamic state measured by both Ag/AgCl electrodes and the MAE are seriously disturbed by motion artifacts, but ECG signals of the MAE is more distinguishable than the signal collected by Ag/AgCl electrodes. Both the MAE and Ag/AgCl electrodes can record EEG signals. The MAE could directly capture bio-signals without skin preparation and application of electrolytic gel.

MAE seems to be a promising alternative electrode to conventional Ag/AgCl electrodes in bio-signal monitoring. The comfort of MAE for patients during the long-term acquisitions and the strength of the micro-needles need to be further improved. Additionally, before the proposed MAE electrode is applied for EEG monitoring in hairy positions, a special fixture with a suitable pressure on the head should be designed and fabricated. Acknowledgments: This research is financially supported by the National Nature Science Foundation of China (Project No. 51575543), Natural Science Foundation of Guangdong Province, China (Project No. 2014A030313211), and Science and Technology Planning Project of Guangdong Province, China (No. 2015B010125004). Author Contributions: Lei Ren designed the setup and wrote the entire manuscript. Qing Jiang and Lelun Jiang managed the research and supervised the revision of the manuscript. Keyun Chen, Zhipeng Chen and Chengfeng Pan participated in the experiments and data processing. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2.

3. 4. 5.

6. 7.

Spinelli, E.; Haberman, M. Insulating electrodes: A review on biopotential front ends for dielectric skin-electrode interfaces. Physiol. Meas. 2010, 31. [CrossRef] [PubMed] Forvi, E.; Bedoni, M.; Carabalona, R.; Soncini, M.; Mazzoleni, P.; Rizzo, F.; O’Mahony, C.; Morasso, C.; Cassarà, D.G.; Gramatica, F. Preliminary technological assessment of microneedles-based dry electrodes for biopotential monitoring in clinical examinations. Sens. Actuators A Phys. 2012, 180, 177–186. [CrossRef] Zhou, W.; Song, R.; Pan, X.; Peng, Y.; Qi, X.; Peng, J.; Hui, K.S.; Hui, K.N. Fabrication and impedance measurement of novel metal dry bioelectrode. Sens. Actuators A Phys. 2013, 201, 127–133. [CrossRef] Li, Z.; Li, Y.; Liu, M.; Cui, L.; Yu, Y. Microneedle electrode array for electrical impedance myography to characterize neurogenic myopathy. Ann. Biomed. Eng. 2015, 44, 1566–1575. [CrossRef] [PubMed] O’Mahony, C.; Pini, F.; Blake, A.; Webster, C.; O’Brien, J.; McCarthy, K.G. Microneedle-based electrodes with integrated through-silicon via for biopotential recording. Sens. Actuators A Phys. 2012, 186, 130–136. [CrossRef] Salvo, P.; Raedt, R.; Carrette, E.; Schaubroeck, D.; Vanfleteren, J.; Cardon, L. A 3D printed dry electrode for ECG EEG recording. Sens. Actuators A Phys. 2012, 174, 96–102. [CrossRef] Yu, L.M.; Tay, F.E.H.; Guo, D.G.; Xu, L.; Yap, K.L. A microfabricated electrode with hollow microneedles for ECG measurement. Sens. Actuators A Phys. 2009, 151, 17–22. [CrossRef]

Sensors 2016, 16, 908

8.

9. 10. 11.

12. 13. 14. 15. 16.

17. 18. 19.

20. 21. 22.

23.

24. 25.

26. 27. 28.

29.

12 of 13

Srivastava, A.K.; Bhartia, B.; Mukhopadhyay, K.; Sharma, A. Long term biopotential recording by body conformable photolithography fabricated low cost polymeric microneedle arrays. Sens. Actuators A Phys. 2015, 236, 164–172. [CrossRef] Hsu, L.S.; Tung, S.W.; Kuo, C.H.; Yang, Y.J. Developing barbed microtip-based electrode arrays for biopotential measurement. Sensors 2014, 14, 12370–12386. [CrossRef] [PubMed] Griss, P.; Enoksson, P.; Tolvanen-Laakso, H.K.; Merilainen, P.; Ollmar, S.; Stemme, G. Micromachined Electrodes for Biopotential Measurements. J. Microelectromech. Syst. 2001, 10, 10–16. [CrossRef] Ng, W.C.; Seet, H.L.; Lee, K.S.; Ning, N.; Tai, W.X.; Sutedja, M.; Fuh, J.Y.H.; Li, X.P. Micro-spike EEG electrode and the vacuum-casting technology for mass production. J. Mater. Process. Technol. 2009, 209, 4434–4438. [CrossRef] Chinteng, L.; Lunde, L.; Yuhang, L.; Lin, B.S.; Chang, J.Y. Novel dry polymer foam electrodes for long-term eeg measurement. IEEE Trans. Biomed. Eng. 2011, 58, 1200–1207. [CrossRef] [PubMed] Griss, P.; Tolvanen-Laakso, H.K.; Merilainen, P.; Stemme, G. Characterization of micromachined spiked biopotential electrodes. IEEE Trans. Biomed. Eng. 2002, 49, 597–604. [CrossRef] [PubMed] Wang, R.X.; Zhao, W.; Wang, W.; Li, Z.H. A Flexible Microneedle Electrode Array With Solid Silicon Needles. J. Microelectromech. Syst. 2012, 21, 1084–1089. [CrossRef] Wilke, N.; Hibert, C.; O’Brien, J.; Morrissey, A. Silicon microneedle electrode array with temperature monitoring for electroporation. Sens. Actuators A. Phys. 2005, 123–124, 319–325. [CrossRef] Xiang, Z.; Yen, S.-C.; Xue, N.; Sun, T.; Tsang, W.M.; Zhang, S.; Liao, L.-D.; Thakor, N.V.; Lee, C. Ultra-thin flexible polyimide neural probe embedded in a dissolvable maltose-coated microneedle. J. Micromech. Microeng. 2014, 24, 065015. [CrossRef] Kim, M.; Kim, T.; Kim, D.S.; Chung, W.K. Curved microneedle array-based semg electrode for robust long-term measurements and high selectivity. Sensors 2015, 15, 16265–16280. [CrossRef] [PubMed] Sabir, M.I.; Xu, X.; Li, L. A review on biodegradable polymeric materials for bone tissue engineering applications. J. Mater. Sci. 2009, 44, 5713–5724. [CrossRef] Xu, S.; Dai, M.; Xu, C.; Chen, C.; Tang, M.; Shi, X.; Dong, X. Performance evaluation of five types of Ag/AgCl bio-electrodes for cerebral electrical impedance tomography. Ann. Biomed. Eng. 2011, 39, 2059–2067. [CrossRef] [PubMed] Lu, T.J.; Xu, F. Mechanical properties of skin: A review. Adv. Mech. 2008, 38, 393–426. Xu, F.; Lu, T.J.; Seffen, K.A. Biothermomechanics of skin tissues. J. Mech. Phys. Solids. 2008, 56, 1852–1884. [CrossRef] Wildnauer, R.H.; Bothwell, J.W.; Douglass, A.B. Stratum corneum biomechanical propepties I. Influence of relative humidity on normal and extracted human stratum corneum. J. Investig. Dermatol. 1971, 56, 72–78. [CrossRef] [PubMed] Hendriks, F.M.; Brokken, D.; Oomens, C.W.; Bader, D.L.; Baaijens, F.P. The relative contributions of different skin layers to the mechanical behavior of human skin in vivo using suction experiments. Med. Eng. Phys. 2006, 28, 259–266. [CrossRef] [PubMed] Gattiker, G.E.; Kaler, K.V.I.S.; Mintchev, M.P. Electronic Mosquito: Designing a semi-invasive Microsystem for blood sampling, analysis and drug delivery applications. Microsyst. Technol. 2005, 12, 44–51. [CrossRef] Elkhyat, A.; Carol, C.M.; Gharbi, T.; Humbert, P. Influence of the hydrophobic and hydrophilic characteristics of sliding and slider surfaces on friction coefficient: In vivo human skin friction comparison. Skin. Res. Technol. 2004, 10, 215–221. [CrossRef] [PubMed] Ulery, B.D.; Nair, L.S.; Laurencin, C.T. Biomedical applications of biodegradable polymers. J. Polym. Sci. Pol. Phys. 2011, 49, 832–864. [CrossRef] [PubMed] Kumari, A.; Yadav, S.K.; Yadav, S.C. Biodegradable polymeric nanoparticles based drug delivery systems. Colloids Surf. B 2010, 75, 1–18. [CrossRef] [PubMed] Lan, S.; Veiseh, M.; Zhang, M. Surface modification of silicon and gold-patterned silicon surfaces for improved biocompatibility and cell patterning selectivity. Biosens. Bioelectron. 2005, 20, 1697–1708. [CrossRef] [PubMed] Zingler, S.; Erber, R.; Lux, C.J.; Seeberger, R.; Bister, D.; Ludwig, B. Biocompatibility of gold and stainless steel chains used for forced eruption of impacted teeth—An in vitro investigation. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. 2013, 116, 159–168. [CrossRef] [PubMed]

Sensors 2016, 16, 908

30.

31. 32.

33.

13 of 13

Larraneta, E.; Moore, J.; Vicente-Perez, E.M.; Gonzalez-Vazquez, P.; Lutton, R.; Woolfson, A.D.; Donnelly, R.F. A proposed model membrane and test method for microneedle insertion studies. Int. J. Pharm. 2014, 472, 65–73. [CrossRef] [PubMed] Mann, D.L.; Zipes, D.P.; Libby, P.; Bonow, R.O. Braunwald’s Heart Disease: A Textbook of Cardiovascular Medicine, 5th ed.; Braunwald, E., Ed.; Philadelphia, W.B. Saunders Co.: St. Louis, MO, USA, 1997. YU, L.M.; Tay, F.E.H.; Guo, D.G.; Xu, L.; Nyan, M.N.; Chong, F.W.; Yap, K.L.; Xu, B. A MEMS-based Bioelectrode for ECG measurement. In Proceedings of the IEEE Sensors 2008 Conference, Lecce, Italy, 26–29 October 2008; pp. 1068–1071. Merletti, R. The electrode-skin interface and optimal detection of bioelectric signals. Physiol. Meas. 2010, 31, 10. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).