micromachines Article
Tongue Pressure Sensing Array Integrated with a System-on-Chip Embedded in a Mandibular Advancement Splint Yun-Ting Chen 1 , Kun-Ying Yeh 2 , Szu-Han Chen 3 , Chuang-Yin Wang 1 , Chao-Chi Yeh 1 , Ming-Xin Xu 1 , Shey-Shi Lu 2 , Yunn-Jy Chen 3 and Yao-Joe Yang 1, * 1
2 3
*
Department of Mechanical Engineering, National Taiwan University, Taipei 10617, Taiwan;
[email protected] (Y.-T.C.);
[email protected] (C.-Y.W.);
[email protected] (C.-C.Y.);
[email protected] (M.-X.X.) Department of Electronics Engineering, National Taiwan University, Taipei 10617, Taiwan;
[email protected] (K.-Y.Y.);
[email protected] (S.-S.L.) Department of Dentistry, National Taiwan University Hospital, Taipei 10048, Taiwan;
[email protected] (S.-H.C.);
[email protected] (Y.-J.C.) Correspondence:
[email protected]; Tel.: +886-2-3366-2712
Received: 6 June 2018; Accepted: 10 July 2018; Published: 14 July 2018
Abstract: Obstructive sleep apnea (OSA), which is caused by obstructions of the upper airway, is a syndrome with rising prevalence. Mandibular advancement splints (MAS) are oral appliances for potential treatment of OSA. This work proposes a highly-sensitive pressure sensing array integrated with a system-on-chip (SoC) embedded in a MAS. The device aims to measure tongue pressure distribution in order to determine the efficacy of the MAS for treating OSA. The flexible sensing array consists of an interdigital electrode pair array assembled with conductive polymer films and an SoC capable of retrieving/storing data during sleep, and transmitting data for analysis after sleep monitoring. The surfaces of the conductive polymer films were patterned with microdomed structures, which effectively increased the sensitivity and reduced the pressure sensing response time. The measured results also show that the crosstalk effect between the sensing elements of the array was negligible. The sensitivity of the sensing array changed minimally after the device was submerged in water for up to 100 h. Keywords: obstructive sleep apnea; mandibular advancement splint; pressure sensing array; conductive polymer; system-on-chip
1. Introduction Epidemiological studies published during the past decade have shown that the prevalence of obstructive sleep apnea (OSA) was, on average, about 22% in men and 17% in women. Some studies have also shown that the prevalence has increased significantly (from 2008 to 2013, a 37% increase for men and a 50% increase for women) [1]. Hypercapnia, hypoxemia, and sleep fragmentation are typical OSA symptoms and are caused by the collapse of the upper airway while sleeping [2]. Patients with OSA are at higher risk of being involved in a vehicle crash due to OSA-related characteristics such as daytime sleepiness, high apnea-hypopnea indices (AHI), and low oxygen saturation levels [3]. Moreover, OSA may increase the possibility of fatal cardiovascular events [4]. Nasal continuous positive airway pressure (CPAP) is the most common treatment for OSA. Although health-related improvements with CPAP have been demonstrated, the inconvenience and discomfort of using CPAP has resulted in relatively low acceptance rate by patients, which limits its clinical effectiveness. In recent years, oral appliances, particularly mandibular advancement Micromachines 2018, 9, 352; doi:10.3390/mi9070352
www.mdpi.com/journal/micromachines
Micromachines 2018, 9, 352
2 of 15
splints (MAS), which increase the airway lumen volume by shifting the positions of the jaw and tongue, have been developed as a potential alternative treatment for OSA [5]. While MAS have been demonstrated to be effective for some patients, they may be ineffective for other patients or even result in worse OSA syndromes. Therefore, various research teams around the world have reported studies on the fundamental mechanism and efficacy of MAS therapy for OSA. In general, the change in the airway lumen volume caused by using MAS can be observed by using tools such as endoscopy [6,7], lateral cephalometric radiography [8,9], fluoroscopy [10,11], computed tomography (CT) [12,13], and magnetic resonance imaging (MRI) [14,15]. In Ref. [6], Ryan et al. studied the correlation between the improvement in OSA symptoms and the changes in pharyngeal dimensions produced by a MAS by videoendoscopy of the upper airway. Mehta et al. [8] conducted a randomized controlled study of MAS for OSA using lateral cephalometric radiography. This study suggests that treatment outcome of MAS or OSA can be predicted by a combination of anthropomorphic, polysomnographic, and radiographic measurements. In Ref. [10], lateral cephalometric radiography and fluoroscopy were demonstrated to be effective at evaluating if the advancement of the mandibular will increase the oropharyngeal airway of a patient. In addition, cine CT was used to investigate the effect of MAS on pharyngeal size and shape [12]. Cross-sectional CT images recorded pharyngeal changes for two cycles of respiration. Sanner et al. [14] presented a method for predicting the efficacy of MAS for OSA by using ultrafast MRI to record the pharynx images for patients during both transnasal shallow respiration and performance of the Muller maneuver. These tools and methods are well-established and capable of acquiring accurate volumetric pharyngeal images. However, some of them are quite expensive or bulky; therefore, the measurements can only be performed in hospitals. As a result, these tools are not suitable for conducting measurement of OSA patients during sleep. This work proposes embedding a tactile sensing array in a MAS to monitor the position and force of a patient’s tongue during sleep. The sensing device, which was fabricated by micromachining techniques, consisted of a patterned conductive polydimethylsiloxane (PDMS) film and a flexible printed circuit board (FPCB) with an array of interdigital electrode pairs. Multi-wall carbon nanotubes (MWCNT) were employed as the conductive fillers in the PDMS film. Also, microdomed structures were transferred onto the surface of the PDMS film using a nylon membrane filter as the mold. The sensor array was connected to a system-on-chip (SoC) capable of retrieving/storing data during sleep, and transmitting data for analysis after sleep monitoring. 2. Device Design and Operational Principles 2.1. Sensor Array Principle and Design During sleep, gravity may contribute to the falling back of the soft palate and tongue; this in turn narrows the upper airway and results in OSA symptoms [16], as shown in Figure 1a,b. Therefore, detecting the tongue force imposed on the palate as well as the position of the tongue during sleep may help improve MAS when treating OSA [17]. In this work, the proposed tactile sensing array embedded in a MAS is capable of measuring these physical quantities. Figure 2a illustrates how the pressure force distribution exerted on the palate during sleep can be measured by the proposed sensing array on the MAS. Figure 2b shows that the falling back of the tongue due to gravity in the supine posture can be readily detected by the sensing array.
Micromachines 2018, 9, 352 Micromachines 2018, 9, x FOR PEER REVIEW Micromachines 2018, 9, x FOR PEER REVIEW
3 of 15 3 of 14 3 of 14
Airflow Airflow Palate Palate Tongue Tongue Airway Airway
(a) (a)
(b) (b)
Figure 1. Illustration of tongue the tongue falling to gravity. (a) Patient’s tongue touches palate Figure 1. Illustration of the falling backback due due to gravity. (a) Patient’s tongue touches ontoonto palate Figure 1. Illustration of the tongue falling back due to gravity. (a) Patient’s tongue touches onto palate when sleeping. (b) Patient’s tongue falls back due to gravity. when sleeping. (b) Patient’s tongue falls back due to gravity. when sleeping. (b) Patient’s tongue falls back due to gravity.
Max Max
MAS MAS Sensing array Sensing array Tongue Tongue
min min (a)
(a)
(b)
(b)
Figure 2. Working principle of device: (a) Tongue pressure distribution measured by the sensing array. Figure 2. Working principle of device: (a) Tongue pressure distribution measured by the sensing Figure 2. Working principle of device: pressure (b) Change of pressure distribution after (a) the Tongue tongue falls back. distribution measured by the sensing array. (b) Change of pressure distribution after the tongue falls back. array. (b) Change of pressure distribution after the tongue falls back.
Figure 3a,b show exploded and schematic views, respectively, of the proposed tongue Figure 3a,b3a,b show exploded andand schematic views, respectively, of the proposed tongue pressureFigure show exploded schematic views, respectively, of the proposed pressurepressure-sensing device. As shown in Figure 3a, the sensing array consisted of four tongue PDMS-MWCNT sensing device. As shown in Figure 3a, 3a, the the sensing array consisted of four PDMS-MWCNT polymer sensing device. As an shown Figure array consisted of four polymer polymer films and FPCBinwith an array ofsensing interdigital electrode pairs (FlexPDMS-MWCNT PCB, Ping-Yi Electronics films andand an an FPCB with an an array of interdigital electrode pairs (Flex PCB, Ping-Yi Electronics films FPCB with array of interdigital electrode pairs (Flex PCB, Ping-Yi Electronics Corporation, Taoyuan, Taiwan). By using a nylon membrane filter as the mold with numerous Corporation, Taoyuan, Taiwan). By By using a nylon membrane filter as the mold with numerous Corporation, Taoyuan, Taiwan). using membrane as the numerous micropores, microdomed structures werea nylon patterned on thefilter surface ofmold the with polymer films. micropores, microdomed structures were patterned on on the the surface of the polymer films. These micropores, microdomed structures were patterned surface polymer films. These These randomly distributed microdomed structures, which have ofanthe average base width of randomly distributed microdomed structures, which have an average basebase width of approximately randomly distributed microdomed structures, which have an average width of approximately approximately 4.5 µm, are essential to achieve high sensitivity for the pressure sensing element. 4.5 μm, are are essential to achieve high sensitivity for for the the pressure sensing element. Figure 3b shows the the 4.5 μm, essential to achieve high sensitivity sensing element. Figure 3b shows Figure 3b shows the assembled device embedded in a pressure MAS. assembled device embedded in a MAS. assembled embedded in a MAS. view of the contacting interface between the polymer film Figuredevice 3c shows a cross-sectional Figure 3c shows a cross-sectional view of the contacting interface between the the polymer filmfilm andand Figure 3c shows a cross-sectional of the contacting between polymer and the interdigital electrode pair of view the FPCB (without theinterface SoC circuit). As uniform pressure is the the interdigital electrode pair of the FPCB (without the the SoCSoC circuit). As uniform pressure is applied, interdigital electrode ofpart the FPCB (without circuit). As uniform pressure is applied, applied, as shown in thepair right of Figure 3c, the microdomed structures are deformed, which as shown in the right partpart of Figure 3c, the microdomed structures are are deformed, which in turn causes as shown in the of Figure microdomed which in turn causes in turn causes aright rapid increase in3c, thethe contact surfacestructures area betweendeformed, the conductive polymer and a rapid increase in the contact surface area between the conductive polymer and the planar electrode. a rapid increase in the contact surface between the conductive theapplied planar electrode. the planar electrode. Therefore, the area contact resistance decreasespolymer sharply and as the pressure Therefore, the the contact resistance decreases sharply as the applied pressure increases. This sharp Therefore, resistance decreases sharply as the applied pressure increases. sharp increases. This contact sharp resistance change induced by small external forces is the so-calledThis tunneling resistance change induced by small external forces is the so-called tunneling piezoresistive effect [18].[18]. resistance change induced by small external forces is the so-called tunneling piezoresistive effect piezoresistive effect [18]. Devices with this type of piezoresistive effect are much more sensitive than Devices with thisthis typetype of piezoresistive effect are are much more sensitive thanthan typical conductive Devices with of piezoresistive much more sensitive typical conductive typical conductive polymer-based devices effect [19]. polymer-based devices [19]. polymer-based devices [19].
Micromachines Micromachines2018, 2018,9, 9,352 x FOR PEER REVIEW Micromachines 2018, 9, x FOR PEER REVIEW
44 of of 15 14 4 of 14
Conductive Conductive polymer polymer
MAS MAS
Applied pressure Applied pressure
Microdome Microdome
FPCB FPCBarray Electrode Electrode array
Rsurf Rsurf
SoC circuit SoC circuit board board (a) (a)
Rsurf Rsurf Rsurf Rsurf
Electrode Electrode
(b) (b)
(c) (c)
Rsurf Rsurf
Contact area Contact area
Figure 3. of the tongue pressure-sensing device: (a) exploded view of view the sensing (b) Figure 3. Diagrams Diagrams tongue pressure-sensing device: (a) exploded of thearray, sensing Figure 3. Diagrams of of thethe tongue pressure-sensing device: (a) exploded view of the sensing array, (b) schematic the deviceof embedded in a mandibular splint advancement (MAS), and (c) cross-sectional array, (b) of the device embedded advancement in a mandibular (MAS), schematic ofschematic the device embedded in a mandibular advancement splint (MAS), and (c) splint cross-sectional view(c)ofcross-sectional the sensing element before and after pressure force is after applied. and view of the sensing element before and pressure force is applied. view of the sensing element before and after pressure force is applied.
2.2. 2.2. SoC Architecture Architecture 2.2. SoC Architecture The thethe sensing array is capable of retrieving the detected signal from thefrom sensing The SoC SoCattached attachedtoto sensing array is capable of retrieving the detected signal the The SoC attached to the sensing array is capable of retrieving the detected signal from the array, wirelessly recharging the Li-ionthe battery, wirelessly transmitting signals. Asignals. block diagram sensing array, wirelessly recharging Li-ionand battery, and wirelessly transmitting A block sensing array, wirelessly recharging the Li-ion battery, and wirelessly transmitting signals. A block of the SoC shown Figurein4.Figure The system of an analog front-end (AFE) circuit, a 10-bita diagram ofisthe SoC isinshown 4. The consists system consists of an analog front-end (AFE) circuit, diagram of the SoC is shown in Figure 4. The system consists of an analog front-end (AFE) circuit, a successive-approximation-register (SAR) analog-to-digital converter converter (ADC), a low-power 10-bit successive-approximation-register (SAR) analog-to-digital (ADC), a transmitter, low-power 10-bit successive-approximation-register (SAR) analog-to-digital converter (ADC), a low-power and a digital processor, which includes awhich digitalincludes micro-controller (MCU) and a power management transmitter, and a digital processor, a digitalunit micro-controller unit (MCU) and a transmitter, and a digital processor, which includes a digital micro-controller unit (MCU) and a power management unit (PMU). unit (PMU). power management unit (PMU). Wireless Wireless Power Power Transmitter Transmitter
Tactile Tactile Sensing Array Sensing Array
AC-DC AC-DC Rectifier Rectifier
Voltage Voltage Limiter Limiter
Li-ion Battery Li-ion Battery Charger Charger ICHG ICHG
Analog Front-end Analog Front-end Interface Circuit Interface Circuit
VBAT VBAT
Li-ion Li-ion Battery Battery
PFM Buck PFM Buck Regulator Regulator
Digital Processor Digital Processor 10Bit 10Bit Successive Successive Approximation Approximation ADC ADC
Ant. Ant.
UART TX UART TX Digital Signal Processing Digital Signal Processing Power Management Power Management
Low Power Low Power MICS Band MICS Band Transmitter Transmitter
Sensor Control Sensor Control System Clock Divider System Clock Divider
Figure 4. 4. System-on-chip (SoC) (SoC) block diagram. diagram. Figure Figure 4. System-on-chip System-on-chip (SoC) block block diagram.
The AFE AFE circuit circuit utilizes utilizes aa low-power low-power operational operational amplifier amplifier to to form form a resistor feedback amplifier The The AFE circuit utilizes a low-power operational amplifier to form a resistor feedback amplifier with aa loop loop gain gain defined defined by by the the ratio ratio of of resistors. resistors. The The SAR SAR ADC ADC is is utilized utilized in in the the proposed proposed system system with with a loop gain defined by the ratio of resistors. The SAR ADC is utilized in the proposed system becauseof ofits itsadvantages advantagessuch such low power, high resolution, topological complexity. The because asas low power, high resolution, andand low low topological complexity. The SAR because of its advantages such as low power, high resolution, and low topological complexity. The SAR ADC is only suitable for operating at a relatively low frequency, which is appropriate in ADC is only suitable for operating at a relatively low frequency, which is appropriate in biomedical SAR ADC is only suitable for operating at a relatively low frequency, which is appropriate in biomedical applications. Theofoperation the is SAR ADC on a binary-search which applications. The operation the SAR of ADC based onisabased binary-search algorithm,algorithm, which finds the biomedical applications. The operation of the SAR ADC is based on a binary-search algorithm, which n sequentially. final by comparing sampled input signal with the voltages of Vthe /2 findsinterval the final interval bythe comparing the sampled input signal with voltages of V / 2nn ref finds the final interval by comparing the sampled input signal with the voltages of Vref / ref 2 sequentially. 3. Fabrication sequentially. 3. Fabrication 3.1. Fabrication of the Conductive Polymer 3. Fabrication
The conductive polymer was a composite of MWCNT and PDMS. The prepolymer for the 3.1. Fabrication of the Conductive Polymer 3.1. Fabrication ofcomposite the Conductive PDMS-MWCNT was Polymer prepared by dispersing MWCNT (Golden Innovation Business Corp., The conductive polymer was a composite of MWCNT and PDMS. The prepolymer for the New The Taipei, Taiwan) with diameters approximately 20 nm and from 10 to for 50 µm conductive polymer was aofcomposite of MWCNT andlengths PDMS.ranging The prepolymer the PDMS-MWCNT composite was prepared by dispersing MWCNT (Golden Innovation Business Corp., into hexane and composite mixing with prepolymer (Sylgard 184A, Dow Innovation Corning Corp., Midland, PDMS-MWCNT wasPDMS prepared by dispersing MWCNT (Golden Business Corp., New Taipei, Taiwan) with diameters of approximately 20 nm and lengths ranging from 10 to 50 μm New Taipei, Taiwan) with diameters of approximately 20 nm and lengths ranging from 10 to 50 μm into hexane and mixing with PDMS prepolymer (Sylgard 184A, Dow Corning Corp., Midland, MI, into hexane and mixing with PDMS prepolymer (Sylgard 184A, Dow Corning Corp., Midland, MI,
Micromachines2018, 2018,9,9,352 x FOR PEER REVIEW Micromachines
14 55ofof15
USA) at a ratio of 4:1. The mixture was blended with a stirrer for 2 h. The concentration of MWCNT MI, USA) at a ratio of 4:1. The mixture was blended with that a stirrer for 2 h.served The concentration of in the PDMS-MWCNT prepolymer was about 6 wt %. Note the hexane as the dispersant MWCNT in the PDMS-MWCNT prepolymer was about 6 wt %. Note that the hexane served as the for improving prepolymer uniformity during blending. Then, the curing agent (Sylgard 184B, Dow dispersant for improving uniformity blending. Then, theprepolymer curing agent Corning Corp., Midland, prepolymer MI, USA) was dispersedduring with the PDMS-MWCNT at (Sylgard a ratio of 184B, Dow Corning Corp., Midland, MI, USA) was dispersed with the PDMS-MWCNT prepolymer at 10:1 and stirred for 50 min. After degassing in a chamber for 120 min, the PDMS-MWCNT aprepolymer ratio of 10:1was andready stirred for 50 min. After degassing in a chamber for 120 min, the PDMS-MWCNT for the subsequent soft-lithography process. prepolymer ready for the Figure was 5 illustrates the subsequent fabricationsoft-lithography process of the process. conductive polymer with microdomed Figure 5 illustrates the fabrication process of the conductive polymer with 2050, microdomed structures. structures. First, a layer of 200-μm thick-film SU-8 photoresist (SU-8 MicroChem Co., First, a layer of MA, 200-µm thick-film SU-8 photoresist (SU-8 2050, MicroChem Co., Westborough, MA, Westborough, USA) was spin-coated onto a silicon handling wafer, as shown in Figure 5a. Then, ® nylon USA) wasmembrane spin-coated onto a silicon handling wafer, as shown in Figure 5a. Then, a nylon a nylon filter (MS membrane filter, Membrane Solutions Corp., Kent,membrane WA, USA) ® nylon membrane filter, Membrane Solutions Corp., Kent, WA, USA) with a pore size of 3 µm filter (MS with a pore size of 3 μm was placed on the top surface of the SU-8 layer, as shown in Figure 5b. Next, was placed on the of top200-μm surface of the SU-8 as shown in Figure Next, a second with layer of a second layer SU-8 was layer, spin-coated (Figure 5c)5b.and patterned a 200-µm dose of 2 (Figure 5d). As shown in 2 SU-8 was spin-coated (Figure 5c) and patterned with a dose of 300 mJ/cm 300 mJ/cm (Figure 5d). As shown in Figure 5e, the SU-8 mold was formed after SU-8 development. Figure the SU-8 mold was formed after SU-8 development. Then,into a layer 200-µm PDMS-MWCNT Then, 5e, a layer of 200-μm PDMS-MWCNT prepolymer was added the of mold (Figure 5f) and cured ◦ prepolymer added the mold (Figure andwith cured 95 C for 5 h. After removing the SU-8 at 95 °C forwas 5 h. Afterinto removing the SU-8 5f) mold a at stripper (Remover PG, MicroChem Co., mold with a stripper (Remover PG, MicroChem Co., Westborough, MA, USA) (Figure 5g), the patterned Westborough, MA, USA) (Figure 5g), the patterned conductive polymer was peeled from the conductive was peeled from membrane (Figure 5h), and thewere patterns on the nylon membrane polymer filter (Figure 5h), and the the patterns on thefilter nylon membrane filter transferred to the membrane wereFigure transferred to the conductive polymer. Figure(SEM) 6a,b show scanning electron conductivefilter polymer. 6a,b show scanning electron microscopy images of the surfaces of microscopy (SEM) images of the surfaces of the nylon membrane filter and the patterned conductive the nylon membrane filter and the patterned conductive polymer, respectively. The conductive polymer, The placed conductive polymer films were then placed to the top The of the electrode array polymerrespectively. films were then to the top of the electrode array on the FPCB. edges of polymer on the FPCB. The edges of polymer films were sealed by using PDMS prepolymer. films were sealed by using PDMS prepolymer.
Spin-coating base SU-8 (a)
Developing SU-8 mold (e)
Placing membrane filter (b)
Filling conductive polymer (f)
Spin-coating second SU-8 (c)
Removing SU-8 (g)
Patterning SU-8 with UV-light (d)
Peeling off conductive polymer (h)
Wafer
SU-8
Patterned SU-8
Membrane filter
Conductive polymer
Figure5.5.The Thefabrication fabricationprocess processof ofthe theconductive conductivepolymer polymerwith withmicrodomed microdomedstructures. structures. Figure
Micromachines 2018, 9, 352 Micromachines Micromachines 2018, 2018, 9, 9, xx FOR FOR PEER PEER REVIEW REVIEW
6 of 15 66 of of 14 14
Figure SEM pictures of the conductive The scale bars Figure Figure 6. 6. SEM SEM pictures pictures of of the the (a) (a) nylon nylon membrane membrane filter filter and and (b) (b) conductive conductive polymer. polymer. The The scale scale bars bars indicate µm. indicate indicate aa length length of of 50 50 μm. μm.
3.2. Assembling Assembling and Packaging Packaging 3.2. 3.2. Assembling and and Packaging Figure 77 shows shows mandibleand and maxillamolds molds on which a MAS was attached. Figureshows 7a shows Figure Figure 7 shows mandible mandible and maxilla maxilla molds on on which which aa MAS MAS was was attached. attached. Figure Figure 7a 7a shows the the the MAS attached tomaxilla the maxilla mold, and Figure 7b shows the attached MAS attached to the mandible MAS attached to the mold, and Figure 7b shows the MAS to the mandible mold. MAS attached to the maxilla mold, and Figure 7b shows the MAS attached to the mandible mold. mold. The proposed(i.e., device sensing (i.e., the sensing arraySoC with the SoC circuit glued board) was glued to the The The proposed proposed device device (i.e., the the sensing array array with with the the ®SoC circuit circuit board) board) was was glued to to the the palatal palatal plate plate palatal plate of a MAS using denture resin (Hygenic Denture Resin, Coltene Corp., Madrid, Spain). of of aa MAS MAS using using denture denture resin resin (Hygenic (Hygenic®® Denture Denture Resin, Resin, Coltene Coltene Corp., Corp., Madrid, Madrid, Spain). Spain). The The flexible flexible The flexible sensing array was assembled with the conductive polymer films with microdomed sensing sensing array array was was assembled assembled with with the the conductive conductive polymer polymer films films with with microdomed microdomed structures. structures. The The structures. The shape of the sensing array allow was designed to allow to the sensing array to fit onto the the shape shape of of the the sensing sensing array array was was designed designed to to allow the the sensing sensing array array to fit fit onto onto the the curved curved roof roof of of the curved roof of the MAS as closely as possible. The sensing array was composed of 16 electrode pairs. MAS MAS as as closely closely as as possible. possible. The The sensing sensing array array was was composed composed of of 16 16 electrode electrode pairs. pairs. Each Each electrode electrode Each electrode finger was 400 µm wide, and the gap between fingers was 200 µm. The total sensing finger was 200 μm. The total sensing area of finger was was 400 400 μm μm wide, wide, and and the the gap gap between between fingers fingers was 200 μm. The total sensing area of the the 2 . The SoC circuit board was connected to the area of the interdigital electrode array was 477 mm 2 interdigital interdigital electrode electrode array array was was 477 477 mm mm2.. The The SoC SoC circuit circuit board board was was connected connected to to the the sensing sensing array array sensing array by soldering a 16-wire flexible flat cable between them. The PDMS prepolymer was by by soldering soldering aa 16-wire 16-wire flexible flexible flat flat cable cable between between them. them. The The PDMS PDMS prepolymer prepolymer was was applied applied to to the the applied to the solder spot for both strength and isolation. The entire system was coated with a 3-µm solder solder spot spot for for both both strength strength and and isolation. isolation. The The entire entire system system was was coated coated with with aa 3-μm 3-μm biocompatible biocompatible biocompatible Parylene-C film (Parylene-C, La Chi Enterprise Corp., Taipei, Taiwan). Parylene-C Parylene-C film film (Parylene-C, (Parylene-C, La La Chi Chi Enterprise Enterprise Corp., Corp., Taipei, Taipei, Taiwan). Taiwan).
(a) (a)
(b) (b)
Figure MAS attached to Figure 7. 7. The The to the the (a) (a) maxilla maxilla mold mold and and (b) (b) mandible mandible mold. mold. Figure 7. The MAS MAS attached attached to the (a) maxilla mold and (b) mandible mold.
Figure Figure 8a 8a shows shows aa posterior posterior view view of of the the maxilla maxilla and and mandible mandible molds molds with with the the proposed proposed device device attached. This picture picture also also shows that the attached. This shows that the flexibility flexibility of of the the proposed proposed sensing sensing array array device device facilitates facilitates its its installation installation onto onto aa customized customized MAS. MAS. Figure Figure 8b 8b shows shows aa side side view view of of the the molds molds with with the the proposed proposed device attached. device attached.
Micromachines 2018, 9, 352
7 of 15
Micromachines 2018, 2018, 9, 9, xx FOR FOR PEER PEER REVIEW REVIEW Micromachines
of 14 14 77 of
(a) (a)
(b) (b)
Figure 8. (a) Posterior view of the maxilla and mandible molds with the proposed device attached. Figure Figure 8. 8. (a) (a) Posterior Posterior view view of of the the maxilla maxilla and and mandible mandiblemolds moldswith withthe theproposed proposeddevice deviceattached. attached. (b) Side view of the molds with the proposed device attached. (b) (b) Side Side view view of of the the molds molds with with the the proposed proposed device attached.
4. Measurements 4. Measurements Measurements 4. The experimental setup for measuring the piezoresistive characteristics and dynamic responses The experimental experimental setup setup for for measuring measuring the the piezoresistive piezoresistive characteristics characteristics and and dynamic dynamic responses responses The of the sensing elements are shown in Figure 9a,b, respectively. As shown in Figure 9a, the force of the sensing sensing elements elements are shown shown in Figure 9a,b, respectively. As As shown shown in in Figure Figure 9a, 9a, the the force force of applied to the sensing element was measured using a force gauge (HF-1, resolution: 1 mN, ALGOL applied to the sensing element was measured using a gauge 1 mN, ALGOL applied to the sensing element was measured using a force gauge (HF-1, resolution: 1 mN, ALGOL Instrument Corp., Taoyuan, Taiwan) attached attached to to aa vertical vertical translation translation stage stage (SR-300C, (SR-300C, displacement displacement Instrument Corp., Corp., Taoyuan, Taoyuan, Taiwan) Instrument resolution: μm, Dispenser Tech Corp., New New Taipei, Taipei, Taiwan). Taiwan). A A circular polymethylmethacrylate polymethylmethacrylate resolution: 111 μm, µm, Dispenser Dispenser Tech Tech Corp., resolution: polymethylmethacrylate (PMMA) rod of 7.4 mm in diameter was glued on the sensing head of the force gauge. The The resistance The resistance resistance (PMMA) rod of 7.4 mm in diameter was glued on the sensing head of the force gauge. of the sensing element was measured using a source meter (Model 2400, Keithley Instruments Corp., of the the sensing sensing element elementwas wasmeasured measuredusing usingaasource sourcemeter meter(Model (Model2400, 2400,Keithley KeithleyInstruments InstrumentsCorp., Corp., of Solon, USA). Figure 9b shows the experimental setup for measuring the dynamic responses of Solon, OH, USA). Figure 9b shows the experimental setup for measuring the dynamic responses of the Solon, USA). Figure 9b shows the experimental setup for measuring the dynamic responses of the sensing element. A piezoelectric actuator (P-621.20L, Physik Instruments Corp., Karlsruhe, sensing element. A piezoelectric actuator (P-621.20L, Physik Instruments Corp., Karlsruhe, Germany) the sensing element. A piezoelectric actuator (P-621.20L, Physik Instruments Corp., Karlsruhe, Germany) was byamplifier power fed amplifier fed with with square wave signals generated generated by aagenerator function was drivenwas by driven adriven powerby with square wave signals generated by a function Germany) aa power amplifier fed square wave signals by function generator (GFG8216A, GW Instek, New Taipei,The Taiwan). The sensing element was periodically periodically (GFG8216A, GW Instek,GW New Taipei, Taiwan). sensingThe element waselement periodically pressed and generator (GFG8216A, Instek, New Taipei, Taiwan). sensing was pressed and released by a PMMA block attached to the head of the actuator. The contact area of the releasedand by areleased PMMA by block attached to the head oftothe Theactuator. contact area the PMMA block pressed a PMMA block attached theactuator. head of the Theof contact area of the PMMA block is 33Amm mm mm. A A laser Doppler Doppler interferometer was employed employed to measure measure the is 3 mm block × 3 mm. laser××Doppler interferometer was employed to measure the displacement of the the PMMA is 33 mm. laser interferometer was to displacement of the actuator. The transient behaviors of the sensing element were retrieved using actuator. The transient behaviors of the sensing element were retrieved using a simple circuit which displacement of the actuator. The transient behaviors of the sensing element were retrieved using aa simple circuit which converted the resistance resistance of into the sensing sensing element into aa voltage voltage output. output. converted the which resistance of the sensing element a voltage output.into simple circuit converted the of the element Force gauge gauge Force computer computer
Translational Translational stage stage
Doppler Doppler interferometer interferometer
Power Power amplifier amplifier
Piezoelectric Piezoelectric actuator actuator
傳印幸福
Function Function generator generator
傳印幸福 Lawrence Yun-Ting Chen
Lawrence Yun-Ting Chen
Sensor Sensor
Source meter meter Source
Signal condition condition circuit circuit Signal (a) (a)
(b) (b)
Figure 9. (a) (a)Experimental Experimental setup for measuring the piezoresistive piezoresistive characteristics of the the sensing sensing Figure 9. setup forfor measuring the piezoresistive characteristics of the sensing element. Figure 9. (a) Experimental setup measuring the characteristics of element. (b) Experimental setup for measuring the dynamic responses of the sensing element. (b) Experimental setup forsetup measuring the dynamic responsesresponses of the sensing element. (b) Experimental for measuring the dynamic of theelement. sensing element.
Sensing elements elements with with electrode electrode finger finger widths widths of of 300 300 μm μm (Device (Device A) A) and and 400 400 μm μm (Device (Device B) B) were were Sensing fabricated and measured. The gap between fingers was 200 μm for both devices. Figure 10 shows the fabricated and measured. The gap between fingers was 200 μm for both devices. Figure 10 shows the results of the pressure-resistance measurement of the sensing elements. The results indicate that the results of the pressure-resistance measurement of the sensing elements. The results indicate that the proposed devices devices exhibited exhibited high high sensitivities sensitivities at at lower lower pressure pressure region region (below (below 10 10 kPa). kPa). It It is is worth worth proposed
Micromachines 2018, 9, 352
8 of 15
Micromachines 2018,elements 9, x FOR PEER Sensing with REVIEW electrode finger widths of 300 µm (Device A) and 400 µm (Device B) were 8 of 14 Micromachines 2018, 9, measured. x FOR PEER The REVIEW fabricated and gap between fingers was 200 µm for both devices. Figure 10 shows8 of 14
noting theofsensitivity of Device Bmeasurement was larger than that of Device A because Device B hadthat a larger the that results the pressure-resistance of the sensing elements. The results indicate noting the sensitivity of Device Bsensitivities wasfinger largerwidth. than that of Device A the because a larger the that proposed devices high at lower pressure region (below 10Device kPa). ItB ishad worth effective electrode area exhibited due to its larger Table 1 shows sensitivities calculated for noting that the sensitivity of Device B was larger than that of Device A because Device B had a larger effective electrode area due widths. to its larger finger width. 1 shows the sensitivities for devices with various finger Figure 11 shows theTable measured hysteresis effects of calculated the fabricated effective electrode area due to its larger finger width. Table 1 shows the sensitivities calculated for devices with various finger widths. Figure 11 shows the measured hysteresis effects of the fabricated sensing element. The external pressure force slowly increased from zero to 70 kPa, and then retreated devices various finger widths. shows thewas measured hysteresis ofand the fabricated sensing element. The pressure force11slowly increased zero to effects 70 then retreated to zero. Aswith shown inexternal the figure, theFigure hysteresis effect notfrom significant. It kPa, is possibly because the sensing element. The external pressure force slowly increased from zero to 70 kPa, and then retreated to zero. As shown in theoffigure, the hysteresis was notcaused significant. is possibly because the piezoresistive behavior the proposed sensoreffect is primarily by theItsharp change in contact to zero. As shown in the figure, the hysteresis effect was not significant. It is possibly because the piezoresistive of thestructures proposed and sensor primarily caused the sharp change in contact area between behavior microdomed theis electrodes, which by predominates the viscoelastic piezoresistive behavior of the proposed sensor is primarily caused by the sharp change in contact area area between microdomed structures and the electrodes, which predominates the viscoelastic response observed in typical polymer-based piezoresistive material. between microdomed structures and the electrodes, which predominates the viscoelastic response response observed in typical polymer-based piezoresistive material. observed in typical polymer-based piezoresistive material. 30
Device-A Device-B Device-A Device-B
1.0
Device-A Device-B Device-A Device-B
1.0 0.8
25 20
0.8 0.6
20 15
R/R0R/R0
Resistance (MΩ)(MΩ) Resistance
30 25
15 10 10 5
0.6 0.4 0.4 0.2 0.2 0.0
5 0 0
0
20
0
40
20
40
60
80
100
Pressure (kPa) 60
80
0.0 0
120
100
0
120
5
10
Pressure (kPa)
5
10
15
20
15
20
Pressure (kPa)
Pressure (kPa)
(a) (a)
(b) (b)
Figure 10. (a) Measured responses of the devices of various widths of electrode fingers. (b) Figure 10.10.(a) responses ofdevices the devices ofwidths various widths fingers. of electrode fingers. (b) Normalized measured responses atoflower pressure region (0–20ofkPa). Figure (a)Measured Measured responses the of various electrode (b) Normalized Normalized responses at lower pressure region (0–20 kPa). measuredmeasured responses at lower pressure region (0–20 kPa). 1.0
1.0 1.0 0.8
0.6 0.4 0.4 0.2 0.2 0.0 0.0
press release press release
0.8 0.6
|ΔR|/R |ΔR|/R
0.8 0.6
|ΔR|/R |ΔR|/R
1.0 0.8
press release press release
0.6 0.4 0.4 0.2 0.2 0.0
0 0
10 10
20 20
30
40
50
Pressure (kPa) 30
40
50
Pressure (kPa)
(a) (a)
60 60
70 70
80 80
0.0 0 0
5 5
Pressure (kPa)
10
15
10
15
Pressure (kPa)
(b) (b)
Figure 11. (a) hysteresis curve of the sensing element. (b) The subset of (a) pressure Figure 11. Measured (a) Measured hysteresis curve of the sensing element. (b) The subset of in (a)lower in lower Figure 11. (a) Measured hysteresis curve of the sensing element. (b) The subset of (a) in lower pressure region. pressure region. region. Table1.1.Sensitivities Sensitivities of the different finger widths. Table thedevices deviceswith with different finger widths. Table 1. Sensitivities of the devices with different finger widths.
DevicesDevices Device AA B Device DeviceDevice B Devices Device A Device B Finger width 300300 400 Finger(μm) width (µm) 400 Finger width (μm) 300 Gap width (µm) 200 400 Gap width (μm) 200200 200 −1 ) 0.177 0.288 Sensitivity (kPa −1 Gap width(kPa (μm) ) 200 200 Sensitivity 0.177 0.288 Sensitivity (kPa−1) 0.177 0.288 Figure 12 shows the measured dynamic responses of the sensing element as it was periodically Figure shows the measured dynamic responses of the sensing it wasthe periodically pressed and12 released by the piezoelectric actuator. The sensor responseelement lagged as behind motion of pressed and released by the piezoelectric actuator. The sensor response lagged behind the motion of the piezoelectric actuator by about 5 ms. Figure 13 depicts the results of repeatability and drift tests the piezoelectric about 5for ms.2500 Figure 13 depicts thesame results of repeatability and tests by measuring theactuator sensingby element cycles using the experimental setup as drift shown in by measuring the sensing element for 2500 cycles using the same experimental setup as shown in
Micromachines 2018, 9, 352
9 of 15
Figure 12 shows the measured dynamic responses of the sensing element as it was periodically pressed and released by the piezoelectric actuator. The sensor response lagged behind the motion of the piezoelectric actuator by about 5 ms. Figure 13 depicts the results of repeatability and drift tests by measuring the sensing element for 2500 cycles using the same experimental setup as shown in Figure 9b driven with a 1-Hz square wave. Figure 13b is the subset snapshot of Figure 13a. The sensing Micromachines 2018, 9, x FOR PEER REVIEW 9 of 14 Micromachines 2018, 9, xreasonably FOR PEER REVIEW 9 of 14 element exhibited good repeatability and was capable of long-term pressure sensing. 5000 5000 4000 4000
3.6 3.6 3.4 3.4
3000 3000
3.2 3.2
2000 2000
3.0 3.0
1000 1000
Voltage (V)(V) Voltage
Displacement (nm) Displacement (nm)
3.8 3.8
Displacement Voltage Displacement Voltage
2.8 2.8 2.6 2.6
0 0 0 0
100 100
200 200
300 300
Time (ms) Time (ms)
400 400
2.4 6002.4 600
500 500
Voltage (V) (V) Voltage
Figure Figure 12. 12. Dynamic Dynamic square square waves waves response response of of the the sensing sensing element. Figure 12. Dynamic square waves response of the sensing element. Voltage Voltage
4.5 4.5 4.0 4.0 3.5 3.5 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 0 0
500 500
1000 1000
Time (s) Time (s)
1500 1500
2000 2000
2500 2500
Voltage (V)(V) Voltage
(a) (a) 4.5 4.5 4.0 4.0 3.5 3.5 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 940 940
Voltage Voltage
942 942
944 944
946 946
948 948
950 950(s) Time Time (s)
952 952
954 954
956 956
958 958
960 960
(b) (b) Figure 13. Repeatability test of the sensing element: (a) 2500 of the sensing element’s dynamic Figure 13. 13. Repeatability Repeatability test test of of the the sensing sensing element: element: (a) (a) 2500 2500 sss of of the the sensing sensing element’s element’s dynamic dynamic Figure responses. (b) A snapshot of (a) between 940 s and 960 s. responses. (b) A snapshot of (a) between 940 s and 960 s. responses. (b) A snapshot of (a) between 940 s and 960 s.
Figure 14 shows the crosstalk effect measured between the sensing elements. An external Figure 14 shows the crosstalk effect measured between the sensing elements. An external pressure force was exerted on point X, and the resistances at points M, G, L, H, D, and F were pressure force was exerted on point X, and the resistances at points M, G, L, H, D, and F were measured separately. The measured results show that the resistance at point X decreased measured separately. The measured results show that the resistance at point X decreased considerably as pressure force was applied to point X, while the resistances of the adjacent points considerably as pressure force was applied to point X, while the resistances of the adjacent points (i.e., M, G, L, H, D, and F) were almost unchanged even when the applied force was raised to (i.e., M, G, L, H, D, and F) were almost unchanged even when the applied force was raised to 125 kPa, which indicates that the crosstalk effect was not discernible for the proposed sensing array.
Micromachines 2018, 9, 352
10 of 15
Figure 14 shows the crosstalk effect measured between the sensing elements. An external pressure force was exerted on point X, and the resistances at points M, G, L, H, D, and F were measured separately. The measured results show that the resistance at point X decreased considerably as pressure force was applied to point X, while the resistances of the adjacent points (i.e., M, G, L, H, D, Micromachines 2018, 9, xunchanged FOR PEER REVIEW 10 of 14 and F) were almost even when the applied force was raised to 125 kPa, which indicates Micromachines 2018, 9, x FOR PEER REVIEW 10 of 14 that the crosstalk effect was not discernible for the proposed sensing array. 1.0 1.0
X XM MG GL LH HD DF F
0.8
｜ΔR｜/R0 ｜ΔR｜/R0
0.8 0.6
D D F G X H F GL XMH L M
0.6 0.4 0.4 0.2 0.2 0.0 0.0 0 0
20
40
20
40
60 60 Pressure
80 80 (kPa)
100 100
120 120
140 140
Pressure (kPa) Figure 14. The measured normalized resistances for crosstalk effect measurements. The inset shows Figure 14.The The measured normalized resistances foron crosstalk effectmeasurements. measurements. The Theinset insetshows shows the position ofmeasured the corresponding sensing element the array. Figure 14. normalized resistances for crosstalk effect the position of the corresponding sensing element on the array. the position of the corresponding sensing element on the array.
Changes in the proposed device’s sensitivity after the sensing array, which was packaged using Changes proposed sensitivity after was packaged using Changesin inthe thesubmerged proposeddevice’s device’s sensitivity after the sensing array, which was packaged using Parylene-C, was in deionized water forthe 50sensing h, 75 array, h, andwhich 100 h were evaluated to Parylene-C, was in deionized water for 50 for h, 75 and75100 were evaluated demonstrate Parylene-C, was submerged in capability. deionized water h, h,hthe and 100 h wereto evaluated demonstrate itssubmerged water-resistant As shown in50h, Figure 15, pressure-resistance curvestoof its water-resistant capability. Assubmerged shown in Figure 15, the pressure-resistance of the sensing demonstrate its water-resistant capability. shown Figureperiods 15, the pressure-resistance curves of the sensing array after being inAswater forin various of curves time were almost thearray same after being submerged in water for various periods of time were almost same as the original one, the sensing array after being submerged water for various periods ofthe time were almost the same as the original one, which indicates that in the changes in sensitivity were negligible. This preliminary which that the changes in sensitivity were negligible. This preliminary finding shows that as the indicates original indicates that the in sensitivity were negligible. This preliminary finding showsone, thatwhich Parylene-C coating can changes provide reasonably effective short-term water-resistant Parylene-C coating provide reasonably capability. finding shows thatcan Parylene-C coating caneffective provideshort-term reasonablywater-resistant effective short-term water-resistant capability. capability. 80
Initial 50hr Initial 75hr 50hr 100hr 75hr 100hr
80
Resistance (kΩ) Resistance (kΩ)
60 60 40 40 20 20 0 0 0 0
50
100
150
200
250
50
100 150 200 250 Pressure (kPa) Pressure (kPa) Figure15. 15.Pressure–resistance Pressure–resistancecurves curvesafter afterthe thepackaged packageddevice devicewas wassubmerged submergedininwater waterfor fordifferent different Figure Figure 15. Pressure–resistance curves after the packaged device was submerged in water for different periods time. periods ofoftime. periods of time.
Figure 16 shows screenshots of the sensing array’s responses to an artificial silicon tongue Figure 16sensing shows screenshots of the array’s to an artificial silicon touching the array installed on asensing MAS that was responses attached between mandible and tongue maxilla touching theinsets sensing arraysub-figures installed onarea the MAS that was attached between mandible and maxilla molds. The in these pressure distributions measured by the sensing array. molds. The insets in these sub-figures are the pressure distributions measured byresults the sensing array. The color-bar legend is from 5 kPa (cyan) to 50 kPa (red). Figure 16a,b show the of when the The color-bar legend is from 5 kPa (cyan) to 50 kPa (red). Figure 16a,b show the results of when the artificial tongue pressed the center and posterior areas of the sensing array. Figure 16c,d show the
Micromachines 2018, 9, 352
11 of 15
Figure 16 shows screenshots of the sensing array’s responses to an artificial silicon tongue touching the sensing array installed on a MAS that was attached between mandible and maxilla molds. The insets in these sub-figures are the pressure distributions measured by the sensing array. The color-bar legend is from 5 kPa (cyan) to 50 kPa (red). Figure 16a,b show the results of when the artificial tongue pressed the center and posterior areas of the sensing array. Figure 16c,d show the results of the tongue pressing the sinistral and dextral areas of the sensing array. Micromachines 2018, 9, x FOR PEER REVIEW 11 of 14
(a)
(b)
(c)
(d)
Figure Figure 16. 16. Screenshots Screenshots of of the the sensing sensing array array measuring measuring the the pressure pressure distribution distributionof of aa silicon silicon tongue. tongue. The tongue is pressed against the center (a), posterior (b), left (c), and right (d) areas the sensing The tongue is pressed against the center (a), posterior (b), left (c), and right (d) areas of theofsensing array. array.
To achieve a sufficiently compact module fitting in the very limited space of a MAS, a specifically To achieve a sufficiently compact module in the very limited space of a MAS, a specifically designed SoC chip was realized by UMC 0.18 fitting µm standard CMOS process (United Microelectronics designed SoC chip was realized by shows UMC 0.18 μm standard CMOS process (UnitedofMicroelectronics Corp., Hsinchu, Taiwan). Figure 17 the measured input-output characteristic AFE by feeding Corp., Hsinchu, Taiwan). Figure 17 shows the measured input-output characteristic AFE by sinusoidal input signals with amplitudes ranging from 50 to 1800 mV. The blue line is theofregression feeding sinusoidal input signals with amplitudes ranging from 50 to 1800 mV. The blue line is the line of small input voltage with a calculated gain of 1.21. The output amplitude starts to be compressed regression line of small input voltage with850 a calculated gain of 1.21.the The output amplitude starts to when the input amplitude reaches around mV. Figure 18 shows measured ADC output codes be compressed when the input amplitude reaches around 850 mV. Figure 18 shows the measured corresponding to different DC input levels from the AFE. The DC input levels from the AFE were 900, ADC output codes corresponding to different input from the AFE. The DC input levels 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1620, andDC 1640 mV. levels The corresponding output codes of ADC in from the AFE were 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1620, and 1640 mV. The Figure 18 presented a quite linear characteristic. The glitches on these curves resulted from supply corresponding output codeswhich of ADC Figure 18 presented quite linear characteristic. The glitches and environmental noises caninbe simply eliminateda using additional off-chip components. on these resulted from supply environmental which bechip simply Figure 19curves is the micrograph of the SoC and implemented for thenoises sensing array.can The area,eliminated including using additional off-chip components. Figure 19 is the micrograph of the SoC implemented the the testing pads, is 3.16 mm × 3.09 mm. Table 2 lists the measured performance summaryfor which sensing array. The chip area, including the testing pads, is 3.16 mm × 3.09 mm. Table 2 lists the concludes the detailed measurement results of all the sub-blocks in this SoC chip. measured performance summary which concludes the detailed measurement results of all the subblocks in this SoC chip.
corresponding output codes of ADC in Figure 18 presented a quite linear characteristic. The glitches on these curves resulted from supply and environmental noises which can be simply eliminated using additional off-chip components. Figure 19 is the micrograph of the SoC implemented for the sensing array. The chip area, including the testing pads, is 3.16 mm × 3.09 mm. Table 2 lists the measured performance summary which concludes the detailed measurement results of all the Micromachines 2018, 9, 352 12 subof 15 blocks in this SoC chip.
Micromachines 2018, 9, x FOR PEER REVIEW
Figure 17. 17. Measured input–output characteristic characteristic of of the the analog analog front-end front-end (AFE). (AFE). Figure Measured input–output
12 of 14 12 of 14
Code DigitalCode Digital
Micromachines 2018, 9, x FOR PEER REVIEW
Time (s)
Time (s)
Figure 18. Measured analog-to-digital converter (ADC) output codes corresponding to different DC Figure 18. Measured analog-to-digital converter (ADC) output codes corresponding to different DCDC Figure Measured converter (ADC) output codes corresponding to different input 18. levels from the analog-to-digital AFE.
input levels from thethe AFE. input levels from AFE.
Figure 19. The micrograph of the System-on-Chip for the sensing array. Table 2. Summary of measured performance of the SoC.
Figure 19. The micrograph of the System-on-Chip for the sensing array. Figure 19. The micrograph of the System-on-Chip for the sensing array. Summary Technology UMC 0.18um CMOS Table 2. Summary of measured performance of the SoC. Chip Area 3.16 × 3.09 mm2 Supply Voltage Summary 1.8 V Technology Analog Front-end UMC 0.18um CMOS SNR 63.1 ×dB Chip Area 3.16 3.09 mm2 GainBWProduct 4.4 MHz
Micromachines 2018, 9, 352
13 of 15
Table 2. Summary of measured performance of the SoC. Summary Technology Chip Area Supply Voltage
UMC 0.18um CMOS 3.16 × 3.09 mm2 1.8 V Analog Front-end
SNR GainBWProduct Linearity Sensitivity Power Consumption
63.1 dB 4.4 MHz ~0.999 14 mV/kPa 6 mW SAR ADC
Resolution Sampling Rate SFDR DNL INL ENOB Power Consumption
10 bit 200 kS/s 74.0
[email protected] kHz +0.127/−0.18 LSB +0.30/−0.14 LSB 9.47
[email protected] kHz 6.24 mW Digital Control Circuits
Clock Rate Power Consumption
UART: 38.4 KHz; Scan rate: 1 Hz 2.53 mW Buck DC-DC Converter
Input voltage Power Consumption
3.6 V 6 mW @ 98 mA, η = 78.3% OOK Transmitter
Carrier Frequency Output Power Power Consumption
403 MHz (MICS) −14 dBm 1 mW Off-chip Components
Li-ion Battery Power Consumption
3.6 V/50 mAh EEPROM: 0.72 mV @ Read; 9 mV @ Write
5. Conclusions This work presented a highly-sensitive pressure sensing array integrated with an SoC circuit embedded in a MAS for measuring tongue pressure distribution. The proposed flexible device consists of an FPCB with an interdigital electrode array with conductive polymer films and an SoC for retrieving, storing, and transmitting data. The pressure sensing array exhibited high sensitivity and rapid response because of the microdomed structures patterned onto the conductive polymer films. The crosstalk effect between adjacent sensing elements was not obvious. In addition, the device displayed effective short-term water-resistant capability. The proposed device can be used to assist physicians to study and improve the efficacy of MAS for OSA. Author Contributions: The sensing array was designed and fabricated by Y.-T.C., C.-C.Y. and M.-X.X. in Y.-J.Y.’s research group. The SoC was designed by K.-Y.Y. and S.-S.L. The measurement setup was designed and implemented by Y.-T.C. and C.-Y.W. The paper was written by Y.-T.C. and Y.-J.Y. The medical oversights are provided by S.-H.C. and Y.-J.C. The principle investigator is Y.-J.Y. Funding: This research was partially funded by the Ministry of Science and Technology, Taiwan grant number 105-2221-E-002-126.
Micromachines 2018, 9, 352
14 of 15
Acknowledgments: The authors are grateful for Fu Tan’s help on the measurement using an interferometer. Conflicts of Interest: The authors declare no conflicts of interest.
References 1. 2.
3. 4.
5. 6.
7. 8.
9. 10.
11.
12. 13.
14.
15.
16.
17.
Franklin, K.A.; Lindberg, E. Obstructive sleep apnea is a common disorder in the population—A review on the epidemiology of sleep apnea. J. Thorac. Dis. 2015, 7, 1311. [PubMed] Epstein, L.J.; Kristo, D.; Strollo, P.J.; Friedman, N.; Malhotra, A.; Patil, S.P.; Ramar, K.; Rogers, R.; Schwab, R.J.; Weaver, E.M.; et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J. Clin. Sleep Med. 2009, 5, 263. [PubMed] Tregear, S.; Reston, J.; Schoelles, K.; Phillips, B. Obstructive sleep apnea and risk of motor vehicle crash: Systematic review and meta-analysis. J. Clin. Sleep Med. 2009, 5, 573. [PubMed] Marin, J.M.; Carrizo, S.J.; Vicente, E.; Agusti, A.G. Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: An observational study. Lancet 2005, 365, 1046–1053. [CrossRef] Schmidt-Nowara, W.; Lowe, A.; Wiegand, L.; Cartwright, R.; Perez-Guerra, F.; Menn, S. Oral appliances for the treatment of snoring and obstructive sleep apnea: A review. Sleep 1995, 18, 501–510. [CrossRef] [PubMed] Ryan, C.F.; Love, L.L.; Peat, D.; Fleetham, J.A.; Lowe, A.A. Mandibular advancement oral appliance therapy for obstructive sleep apnoea: Effect on awake calibre of the velopharynx. Thorax 1999, 54, 972–977. [CrossRef] [PubMed] Schessl, J.; Rose, E.; Korinthenberg, R.; Henschen, M. Severe obstructive sleep apnea alleviated by oral appliance in a three-year-old Boy. Respiration 2008, 76, 112–116. [CrossRef] [PubMed] Mehta, A.; Qian, J.; Petocz, P.; Darendeliler, M.A.; Cistulli, P.A. A Randomized, controlled study of a mandibular advancement splint for obstructive sleep apnea. Am. J. Respir. Crit. Care Med. 2001, 163, 1457–1461. [CrossRef] [PubMed] Robertson, C.; Herbison, P.; Harkness, M. Dental and occlusal changes during mandibular advancement splint therapy in sleep disordered patients. Eur. J. Orthod. 2003, 25, 371–376. [CrossRef] [PubMed] Battagel, J.M.; L’Estrange, P.R.; Nolan, P.; Harkness, B. The role of lateral cephalometric radiography and fluoroscopy in assessing mandibular advancement in sleep-related disorders. Eur. J. Orthod. 1998, 20, 121–132. [CrossRef] [PubMed] Lee, C.H.; Kim, J.W.; Lee, H.J.; Seo, B.S.; Yun, P.Y.; Kim, D.Y.; Yoon, I.Y.; Rhee, C.S.; Park, J.W.; Mo, J.H. Determinants of treatment outcome after use of the mandibular advancement device in patients with obstructive sleep apnea. Arch. Otolaryngol. Head. Neck. Surg. 2010, 136, 677–681. [CrossRef] [PubMed] Kyung, S.H.; Park, Y.C.; Pae, E.K. Obstructive sleep apnea patients with the oral appliance experience pharyngeal size and shape changes in three dimensions. Angle Orthod. 2005, 75, 15–22. [PubMed] Haskell, J.A.; McCrillis, J.; Haskell, B.S.; Scheetz, J.P.; Scarfe, W.C.; Farman, A.G. Effects of Mandibular Advancement Device (MAD) on Airway Dimensions Assessed with Cone-beam Computed Tomography. In Seminars in Orthodontics; Elsevier: New York, NY, USA, 2009; Volume 15, pp. 132–158. Sanner, B.M.; Heise, M.; Knoben, B.; Machnick, M.; Laufer, U.; Kikuth, R.; Zidek, W.; Hellmich, B. MRI of the pharynx and treatment efficacy of a mandibular advancement device in obstructive sleep apnoea syndrome. Eur. J. Orthod. 2002, 20, 143–150. [CrossRef] Chan, A.S.; Sutherland, K.; Schwab, R.J.; Zeng, B.; Petocz, P.; Lee, R.W.; Darendeliler, M.A.; Cistulli, P.A. The effect of mandibular advancement on upper airway structure in obstructive sleep apnoea. Thorax 2010, 65, 726–732. [CrossRef] [PubMed] Yildirim, N.; Fitzpatrick, M.F.; Whyte, K.F.; Jalleh, R.; Wightman, A.J.; Douglas, N.J. The Effect of posture on upper airway dimensions in normal subjects and in patients with the sleep apnea/hypopnea syndrome. Am. Rev. Respir. Dis. 1991, 144, 845–847. [CrossRef] [PubMed] Chen, Y.T.; Yeh, C.C.; Chen, S.H.; Xu, M.X.; Tan, F.; Chen, Y.J.; Yang, Y.J. A Mandibular Advancement Device Embedded with Polymer-based Tunneling Piezoresistive Pressure Sensing Array. In Proceedings of the 31st IEEE International Conference on Micro Electro Mechanical Systems (MEMS), Belfast, UK, 21–25 January 2018; pp. 51–54.
Micromachines 2018, 9, 352
18.
19.
15 of 15
Park, J.; Lee, Y.; Hong, J.; Ha, M.; Jung, Y.-D.; Lim, H.; Kim, S.Y.; Ko, H. Giant tunneling piezoresistance of composite elastomers with interlocked microdome arrays for ultrasensitive and multimodal electronic skins. ACS Nano 2014, 8, 4689–4697. [CrossRef] [PubMed] Ma, C.W.; Chang, C.M.; Lin, T.H.; Yang, Y.J. Highly sensitive tactile sensing array realized using a novel fabrication process with membrane filters. J. Microelectromech. Syst. 2015, 24, 2062–2070. [CrossRef] © 2018 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/).
