Soft wearable contact lens sensor for continuous intraocular pressure monitoring

Medical Engineering & Physics 36 (2014) 1134–1139

Contents lists available at ScienceDirect

Medical Engineering & Physics journal homepage: www.elsevier.com/locate/medengphy

Soft wearable contact lens sensor for continuous intraocular pressure monitoring Guo-Zhen Chen, Ion-Seng Chan, Leo K.K. Leung, David C.C. Lam ∗ The Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Hong Kong

a r t i c l e

i n f o

Article history: Received 26 November 2013 Received in revised form 11 June 2014 Accepted 16 June 2014 Keywords: Glaucoma Intraocular pressure Contact lens sensor Resonance circuit

a b s t r a c t Intraocular pressure (IOP) is a primary indicator of glaucoma, but measurements from a single visit to the clinic miss the peak IOP that may occur at night during sleep. A soft chipless contact lens sensor that allows the IOP to be monitored throughout the day and at night is developed in this study. A resonance circuit composed of a thin film capacitor coupled with a sensing coil that can sense corneal curvature deformation is designed, fabricated and embedded into a soft contact lens. The resonance frequency of the sensor is designed to vary with the lens curvature as it changes with the IOP. The frequency responses and the ability of the sensor to track IOP cycles were tested using a silicone rubber model eye. The results showed that the sensor has excellent linearity with a frequency response of ∼8 kHz/mmHg, and the sensor can accurately track fluctuating IOP. These results showed that the chipless contact lens sensor can potentially be used to monitor IOP to improve diagnosis accuracy and treatment of glaucoma. © 2014 IPEM. Published by Elsevier Ltd. All rights reserved.

1. Introduction Intraocular pressure (IOP) generally measured during office hours is used as a primary indicator of glaucoma [1]. However, peak IOP occurs at night during sleep [2], and IOP measured during a single visit during office hours would miss the peak. Hughes et al. [3] examined the use of 24-h IOP profile in glaucoma treatment. They found that the use of 24-h IOP profile is useful in facilitating early detection of glaucoma which changed glaucoma treatment in 79% of the cases. IOP profile can be measured using telemetric pressure sensors [4–7] implanted invasively into the eye. Alternately, periodic IOPs can be measured in a hospital sleep lab, where the patient is waked periodically for IOP measurement using standard tonometry. In addition to applanation, IOP can also be determined non-invasively by measuring the corneal curvature using Contact Lens Tonometry (CLT) [8–10]. In these approaches, the local strain change in the contact lens with IOP is sensed by the piezoresistive sensing elements embedded in the sensor. However, the local deformation on the cornea is sensed by the rigid metallic sensing element only after it is mechanically attenuated by the soft lens. After the attenuation,

∗ Corresponding author at: The Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, NT, Hong Kong. Tel.: +852 2358 7208; fax: +852 2358 1543. E-mail addresses: [email protected] (G.-Z. Chen), [email protected] (I.-S. Chan), [email protected] (L.K.K. Leung), [email protected] (D.C.C. Lam). http://dx.doi.org/10.1016/j.medengphy.2014.06.005 1350-4533/© 2014 IPEM. Published by Elsevier Ltd. All rights reserved.

the resultant strain in the embedded sensing element is inherently small. A silicon chip is typically integrated as part of the sensor to amplify and transmit the signal wirelessly. With an encapsulated chip, sensors with embedded chip can be as thick as 583 ␮m [11], while commercial contact lenses for vision correction can be as thin as 50–100 ␮m. The added thickness of the chip based sensor is needed for proper encapsulation of the chip, but the added thickness also would increase the mechanical attenuation and degrade the strain sensitivity of the sensor. Alternate designs using flexible non-metallic conductive film elements to increase the sensitivity has been proposed [12], but the thickness remains relatively thick since silicon chip is still required. To eliminate the silicon chip from the sensor, position-sensitive dyed glycerol-filled micro-chamber was proposed as sensing elements [13]. Instead of sensing local strain changes, the dye is designed to move with the global changes in curvature. The corneal curvature can be determined by tracking the dye movement using a camera. The camera requires a clear line of sight to the eye, which makes it unsuited for use with closed eye during sleep when the peak IOP occurs. Alternately, the change in corneal curvature as a function of IOP can be sensed by an inductive coil (Fig. 1). Changes in the corneal curvature change the inductance of the embedded coils. By coupling the inductive coil (L) with a capacitor (C) to form a resonator, changes in the coil inductance can be detected wirelessly by an external reader as changes in the LC resonance [14]. By establishing the correlation between LC resonance and IOP, the changes in IOP can be read and monitored dynamically by a wireless reader mounted near the eye.

G.-Z. Chen et al. / Medical Engineering & Physics 36 (2014) 1134–1139

1135

Fig. 1. Sensing mechanism of the contact lens sensor for continuous IOP monitoring. (a) Contact lens sensor configuration on an eye at low IOP. (b) Contact lens sensor configuration on an eye at high IOP.

where davg = =

din + dout , 2

(2)

dout − din din + dout

(3)

is the fill ratio of the coil, dout is the outer coil diameter, din is the inner coil diameter,  is the magnetic permeability and n is the number of turns of the coil. The change of the coil diameter davg is, davg (ro + r) · sin (˛ · ro /(2 · (ro + r))) = − 1, davg ro · sin (˛/2)

(4)

where ro is the initial radius of curvature of the contact lens, and ˛ is the opening angle of the circumferential inductive coil, r is the change of the radius of curvature which changes with IOP. For r  ro , the expression can be simplified to, Fig. 2. Structure of the contact lens sensor with an embedded LC resonance circuit. (a) 3-D view of the contact lens sensor. (b) Cross-section view of the contact lens sensor.

davg = a · r, davg

(5)

where In this study, the feasibility of this IOP sensing concept is explored. A soft contact lens sensor with embedded LC resonator is designed and fabricated. The contact lens sensor is tested on silicone rubber model eye and its performance in the human IOP range is examined.

a=

1 ˛ − cot ro 2ro

˛ 2

.

(6)

In earlier studies [16], investigators reported that an IOP change of 1 mmHg generates a change of ∼3 ␮m change in corneal radius (for an eye with its corneal radius of curvature as 7.8 mm) such that the change in IOP can be modeled as, r , c

2. Design and fabrication

IOP =

2.1. Design of the contact lens sensor

where c is dependent on the biomechanical properties of the eye. From Eqs. (5) and (7), it can be determined that the change in coil diameter due to human IOP change is typically less than 0.15 mm, and is small. The change in LS as a function of davg is shown in Fig. 4. The plot showed that the dependence between 11 mm and

Elevated IOP will change the corneal curvature [15–17], and the curvature of the soft contact lens on top. A soft deformable LC resonance circuit (Fig. 2) embedded inside the lens was designed to track the change of the corneal curvature with IOP (Fig. 3). The inductance LS can be calculated using a simple planar spiral coil model with coil spacing s and average coil diameter davg and s <
 2.46  

92.4 92.2 9



+ 0.2 × 2 ,

(1)

Inductance (nH)



1 LS = n2 davg ln 2

(7)

92 91.8 9 91.6 9 91.4 9

Calculate ed inductanc ce

91.2 9 11 Fig. 3. Schematic of the change in inductive coil diameter with contact lens curvature change. The change is exaggerated for clarity, while the typical contact lens curvature is ∼8 mm.

11.07 11.13 Ave rage coil diameter (mm)

11.2

Fig. 4. The inductance of the coil as a linear function of the average coil diameter.

1136

G.-Z. Chen et al. / Medical Engineering & Physics 36 (2014) 1134–1139

Fig. 5. Designed parameters of the inductive coil. (The unit used in the figure is mm.)

11.16 mm is approximately linear such that the change in LS with davg can be approximated as a linear function, LS = m · davg ,

(8)

Fig. 6. Schematic of the fabrication process for the thin film capacitor.

(9)

glass substrate for releasing the inductive coil followed by sputtering (ARC-12M sputtering system) of a 500 nm copper seed layer onto the SU-8. A 30 ␮m resin (AZ9260) was spin-coated onto the seed layer and was lithographically patterned. A 10 ␮m copper was then electoplated onto the exposed copper seed layer. The resin was then removed followed by a dip into copper etching solution (FeCl3 ) to remove the thin 500 nm seed layer. Afterwards, the inductive coil was released from the glass substrate in an ultrasonic bath using a 2 g/L sodium hydroxide solution. The thin film capacitor was fabricated by spin-coating a 10 ␮m SU-8 photo-resist onto a cleaned glass substrate as the sacrificial layer. Then, a 20 nm titanium adhesion layer was sputtered (ARC12M sputtering system) followed by a 1 ␮m gold layer (Fig. 6(a)). Biocompatible phthalocyanine (CuPc) [19,20] was then spin-coated onto the gold layer as the dielectric (Fig. 6(b)). A layer of titanium followed by gold was further sputtered onto the dielectric and patterned using standard metal etching techniques to form the top capacitor electrodes (Fig. 6(c)). Residual photo-resist was then stripped in the MS2001 photoresist stripper bath (Fig. 6(d)) and cut into 3 mm by 1.2 mm thin film capacitors (Fig. 6(e)). Afterwards, the thin film capacitor was soldered to the inductive coil using leadfree solder (Sn96/Ag4) to form the resonance circuit. To ensure biocompatibility and to encapsulate the resonance circuit, the completed resonance circuit was sealed by a conformal coating of 2 ␮m Parylene C (poly-chloro-p-xylylene) (Jiangnan Fine Chemical Co., Ltd., China). Parylene C is biocompatible (Food and Drug Administration approved United States Pharmacopeia Class VI grade), mechanically flexible, and can coat and seal objects conformally [21]. Parylene coating was performed in a Cookson Electronics PDS 2010 system (Specialty Coating Systems Inc., Indianapolis, IN, USA). The resonance circuit was then sandwiched between two medical grade silicone rubber (NuSil MED-6015, Liquid silicone elastomer, NuSil Technology LLC, Carpinteria, CA, USA) layers using mold transfer technology [22,23] (Fig. 7). Contact lens sensor with ∼8.0 mm radius of curvature was fabricated for testing. The design parameters of the contact lens sensor are listed in Table 1.

where m is a constant defined as, m = n2

[ln (2.46/) + 0.2 × 2 ] . 2

Combining Eqs. (8) and (5), the change in inductance is then, LS = m · a · r · davg .

(10)

Further combining Eqs. (10) and (7), the change in inductance with IOP is then, LS = c · m · a · davg · IOP.

(11)

When this inductive coil is coupled with a thin film capacitor, the resonance frequency of the LC sensor can be determined from, fc =

1

2



LS C

,

(12)

where C is the capacitance of the thin film capacitor, such that the change in inductance with frequency is, f LS = −2 . LS fc

(13)

Finally, by combining the Eqs. (11)–(13), the change of IOP as a function of the change in resonance frequency is then, IOP = −

1 2

2ˇ · 2 · a · fc · C · davg

·

f , fc

(14)

where ˇ =c·m

(15)

is an experimentally-determined system constant. Eq. (14) shows that the change in IOP varies linearly with the change in frequency. 2.2. Fabrication of contact lens sensor The sensing elements embedded in the contact lens were fabricated separately. The inductive coil was fabricated independently using standard etching technique. The designed parameters of the inductive coil are shown in Fig. 5. In the fabrication procedure, a 10 ␮m SU-8 photo-resist sacrificial layer was spin coated on a clean

G.-Z. Chen et al. / Medical Engineering & Physics 36 (2014) 1134–1139

1137

Fig. 7. The contact lens sensor with sensing elements embedded in a silicone rubber contact lens.

Fig. 9. Sensor test setup with silicone rubber model eye. Pressure in the eye was generated by injecting water into the corneal chamber through a buret. A needle pressure sensor was inserted into the chamber to monitor the absolute pressure. The resonance frequency of the sensor was monitored wirelessly using a reading coil connected to a network analyzer.

fixture (Fig. 8). The change of silicone eye curvature as a function of the IOP is given by Berest [24] as, R = P

Fig. 8. The silicone rubber model eye for testing the contact lens sensor.

2.3. Fabrication of silicone rubber model eye A silicone rubber model eye with dimensions detailed in Table 2 was prepared for the testing of the contact lens sensor. The model eye was fabricated by mixing 10 parts of the silicone base (KE1310ST; Shin-Etsu Chemical Co., Ltd., Tokyo, Japan) with one part of curing agent (CAT1310; Shi-Etsu Chemical Co., Ltd.). The mixture was cast into a mold and cured at 60 ◦ C for 1 h. After curing, the silicone eye was demolded and mounted onto a custom

Table 1 Inductive contact lens sensor design parameters. Design diameter of the contact lens (mm) Radius of curvature of the contact lens (mm) Central thickness of the contact lens (␮m) Dynamic range (mmHg) Outside diameter of the sensing coil (mm) Inside diameter of the sensing coil (mm) Inductance (nH) Capacitance (pF) Resonance frequency (MHz)

14 8.0 <150 5–40 12.0 10.0 ∼91.1 6.0 ∼215

Table 2 Design parameters of the silicone rubber model eye. Radius of curvature (mm) Thickness of the rubbery cornea (␮m) Poisson ratio Elastic modulus of the silicone rubber (MPa)

7.8 320 0.3 1.5a

a The elastic modulus of the rubber used in this study was experimentally determined from the load–displacement data (F, ıi ) from three-point bending tests (MTS 642.001) [25].

R02 (1 − ) 2Et

,

(16)

where P is the pressure inside the silicone rubber eye chamber with respect to ambient pressure, R0 is the initial radius of curvature of the silicone cornea (P = 0 mmHg), ␯ is the Poisson ratio, t is the central thickness of the silicone rubber cornea, and E is the elastic modulus of the silicone rubber. The silicone cornea was designed to have essentially uniform thickness from the center to the edge. In the tests, water was injected into the silicone model eye via a needle connected to a buret via tubing. The IOP was controlled between 5 mmHg and 45 mmHg in the tests by controlling the water height in the buret, as shown in Fig. 9. 3. Results 3.1. Static tests The contact lens sensor was tested using a silicone rubber model eye to characterize its electrical and IOP sensing performance. The curvature change as a function of pressure was determined using image analysis for pressures set between 5 mmHg and 45 mmHg. Three images were captured using a digital camera (Alpha DSLRA900, 24.6 mega pixel resolution, Sony, Tokyo, Japan) for analysis. The correlation plotted in Fig. 10 showed that the radius of curvature changed at a rate of 3.8 ␮m per mmHg of IOP. The resonance frequency of the sensor was measured wirelessly. A 50-mm-diameter PCB coil was connected to a network analyzer (E5071B, Agilent Technologies Inc., Santa Clara, CA, USA) with standard settings to serve as the reader. Contact lens sensor with radius of curvature of 8.0 mm was mounted onto the silicone rubber model eye. The sensor and the model eye were irradiated with oxygen (O2 ) plasma (Branson/IPC 3000 Plasma Asher, Allwin21 Corp.) for 90 s immediately prior to the measurements [26] to ensure both surfaces were hydrophilic. The contact angle of the silicone increased from 41.2◦ to 120.1◦ after the plasma treatment. The hydrophilic nature of samples immersed and stored in buffered saline can be maintained for 48 h. The experimental frequency responses of the contact lens sensor are shown in Fig. 11. The results showed that the sensing

1138

G.-Z. Chen et al. / Medical Engineering & Physics 36 (2014) 1134–1139

35

Pressure sensor Contact lens

Measured value

7.96

30

Pressure (mmHg)

Radius of Curvature (mm)

7.98

7.94 7.92 7.9 7.88

25 20 15 10

7.86 5

7.84

0

5

10

15

20

25

30

35

Pressure (mmHg)

Resonance Frequency (MHz)

measured data 205 205.05 205.1 205.15 205.2

20

25

30

35

40

45

Applied pressure (mmHg) Fig. 11. Experimental frequency response of the contact lens sensor on silicone rubber model eye.

response varied linearly with IOP (R = 0.998) with a frequency response of ∼8 kHz/mmHg. By fitting of Eq. (14) onto the experimental data in Fig. 12, the system constant ˇ was determined to be −2.62 × 10−9 .

-0.002

Frequency change ( Δ f / f)

150

3.2. Cyclical tests The human IOP is dynamic and is cyclical. The ability of the new sensor to track the dynamic changes of IOP was tested. A high precision needle fiber pressure sensor (OpSens Co., Ltd., Canada) was inserted into the eye chamber to measure the IOP. The ability of the sensor to track the chamber pressure is shown in Fig. 13. The data showed that a pressure lag was present only in the initial cycle, but the lag disappeared after the system stabilized in the second and third cycle. This showed that the thin contact lens sensor can closely and rapidly track the chamber pressure after stabilization. In actual wear, the pressure variation in a day is typically less than 10 mmHg, and fluctuation in IOP can be easily tracked by the contact lens sensor.

204.95

15

100

Fig. 13. Sensor response to IOP change in cyclical test.

Fig. 10. Measured curvature change of the silicone rubber model eye as a function of pressure inside the chamber.

205.25 10

50

Tim e (secon ds)

40

EXP. Eqn. 14 -0.0015

4. Discussions In the design of strain gauge sensing elements with embedded silicon chip developed by Hubanova et al. [11], the central thickness of the contact lens is reported to be 585 ␮m. The thickness is ∼6 times of the commercial contact lens. Because of its large thickness, the sensor is more susceptible to influence from blinking and may lose contact with the cornea, which may result in large error (up to 17 mmHg) in IOP sensing [27]. In this investigation, the chipless sensor thickness is <150 ␮m. Since the noise and reading error of the contact lens sensor is principally from the eye movements and blinking, a thin contact lens would have less noise and error. The tracking performance of the thin contact lens sensor in cyclical tests is shown in Fig. 13. The results showed that changes in IOP can be easily tracked by the new thin contact lens sensor. Moreover, thinner lens would have better oxygen permeability and is more suitable for extended wear. 5. Remarks

-0.001

-0.0005

0 0

10

20

30

40

Pressure difference (mmHg) Fig. 12. Experimental relation between IOP of the silicone eye and frequency change of the sensor.

In this study, a silicone cornea with a radius of curvature of 7.8 mm and a contact lens sensor with a radius of curvature of 8.0 mm were fabricated and tested. The results showed that the contact lens sensor can track the curvature change of the silicone cornea as a function of the IOP, and the resonance frequency of the sensor as a function of the IOP of the silicone eye was shown to be linear and reproducible. However, a single sensor may not fit all eyes. To optimize the IOP tracking performance, sensors with radius of curvature made in increment of 0.25 mm will be fabricated in future studies so that corneas with radius between 7 mm and 8.5 mm can be better fitted to assure good fitting and tracking of IOP in on-going tests on porcine and human eyes.

G.-Z. Chen et al. / Medical Engineering & Physics 36 (2014) 1134–1139

In general, corneal curvature varies between individuals and that the system constant ˇ is dependent on c and m. c is a parameter that characterizes the relation between the corneal curvature change and the IOP; and m is a parameter that characterizes the relation between the frequency and the IOP. In future studies, studies on the influence of individual variations and ˇ are needed to check the general applicability of the relation between IOP and eyes with different curvatures. Preliminary tests showed that the ˇ correlation is applicable to the tracking of IOP between 10 mmHg and 35 mmHg in porcine eyes. 6. Conclusions A chipless inductive coil-based contact lens sensor that tracks contact lens curvature change with IOP was successfully developed and tested. The sensing performance was demonstrated using an external reading coil and a silicone rubber model eye. The results showed that the sensor had an IOP response of ∼8 kHz/mmHg with excellent linearity (R = 0.998). Test results and analysis with theory showed that the sensor can conformally track the pressure fluctuation of the silicone model eye. With good IOP sensing capability, the new sensor can be potentially used for continuous IOP monitoring. Funding None. Ethical approval Not required. Acknowledgements The authors would like to thank the Nanoelectronic Fabrication Facility of the Hong Kong University of Science and Technology for providing supports in fabricating the devices. Conflict of interest Provisional application for US patent on the described new contact lens sensor was applied (US 61/887,452). References [1] Liu JH, Zhang X, Kripke DF, Weinreb RN. “Twenty-four-hour intraocular pressure pattern associated with early glaucomatous changes”. Investig Ophthalmol Vis Sci 2003;44:1586–90. [2] Mosaed S, et al. Correlation between office and peak nocturnal intraocular pressures in healthy subjects and glaucoma patients. Am J Ophthalmol 2005;139(February):320–4.

1139

[3] Hughes E, et al. 24-hour monitoring of intraocular pressure in glaucoma management: a retrospective review. J Glaucoma 2003;12(June):232–6. [4] Collins CC. Miniature passive pressure transensor for implanting in the eye. IEEE Trans Biomed Eng 1967:74–83. [5] Chen PJ, et al. Wireless intraocular pressure sensing using microfabricated minimally invasive flexible-coiled LC sensor implant. J Microelectromech Syst 2010;19:721–34. [6] Chow EY, et al. A miniature-implantable rf-wireless active glaucoma intraocular pressure monitor. IEEE Trans Biomed Circ Syst 2010;4:340–9. [7] Chitnis G, et al. A minimally invasive implantable wireless pressure sensor for continuous IOP monitoring. IEEE Trans Biomed Eng 2013;60: 250–6. [8] Leonardi M, et al. A soft contact lens with a MEMS strain gage embedded for intraocular pressure monitoring. In: Transducers. 12th International Conference on Solid-State Sensors, Actuators and Microsystems; 2003. p. 1043–6. [9] Leonardi M, et al. Wireless contact lens sensor for intraocular pressure monitoring: assessment on enucleated pig eyes. Acta Ophthalmol 2009;87(June):433–7. [10] Mansouri K, Shaarawy T. Continuous intraocular pressure monitoring with a wireless ocular telemetry sensor: initial clinical experience in patients with open angle glaucoma. Br J Ophthalmol 2011;95:627–9. [11] Hubanova R, Aptel F, Chiquet C, Mottet B, Romanet JP. “Effect of overnight wear of the Triggerfish® sensor on corneal thickness measured by Visante® anterior segment optical coherence tomography”. Acta Ophthalmol 2014;92:119–23. [12] Laukhin V, et al. Non-invasive intraocular pressure monitoring with a contact lens engineered with a nanostructured polymeric sensing film. Sens Actuators A: Phys 2011;170:36–43. [13] Yan J. An unpowered, wireless contact lens pressure sensor for pointof-care glaucoma diagnosis. In: Engineering in Medicine and Biology Society, EMBC, 2011 Annual International Conference of the IEEE. 2011. p. 2522–5. [14] Nopper R, et al. Wireless readout of passive LC sensors. IEEE Trans Instrum Meas 2010;59:2450–7. [15] Kasprzak H, et al. Holographic measurement of changes of the central corneal curvature due to intraocular-pressure differences. Opt Eng 1994;33(January):198–203. [16] Lam AKC, Douthwaite WA. The effect of an artificially elevated intraocular pressure on the central corneal curvature. Ophthalmic Physiol Opt 1997;17(January):18–24. [17] Sabry N, et al. Corneal curvature in normal and raised intraocular pressure. Bull Ophthalmol Soc Egypt 1976;69:203–7. [18] Mohan SS, et al. Simple accurate expressions for planar spiral inductances. IEEE J Solid-State Circ 1999;34:1419–24. [19] Maggioni G, et al. Optical response of plasma-deposited zinc phthalocyanine films to volatile organic compounds. Sens Actuators B: Chem 2007;127(October 20):150–6. [20] Zhang QM, et al. An all-organic composite actuator material with a high dielectric constant. Nature 2002;419(September 19):284–7. [21] Noordegraaf J. Conformal coating using parylene polymers. Med Device Technol 1997;8(January–February):14–20. [22] Ho H, et al. Contact lens with integrated inorganic semiconductor devices. In: IEEE 21st International Conference on Micro Electro Mechanical Systems. 2008. p. 403–6. [23] Thomas N, et al. A contact lens with an integrated lactate sensor. Sens Actuators B: Chem 2012;162:128–34. [24] Berest P. Response of a hollow sphere in an elastoviscoplastic medium under a variable internal-pressure. Eur J Mech A: Solids 1990;9:87–102. [25] Young WC. Roark’s formulas for stress and strain. New York: McGraw-Hill Book Co.; 1989. [26] Lai J, et al. Surface modification of silicone rubber by gas plasma treatment. J Adhes Sci Technol 1996;10:231–42. [27] Leonardi M. Microfabricated thin film strain gage sensor with telemetry microprocessor embedded in a soft contact lens for minimally invasive intraocular pressure monitoring; 2007.