Chapter 11: Microelectromechanical systems (MEMS) for in vivo applications – Implantable Sensor Systems for Medical Applications

11

Microelectromechanical systems (MEMS) for in vivo applications

A. Vasudev and S. Bhansali,     Florida International University, USA

Abstract:

This chapter discusses the application of microelectromechanical systems (MEMS) technology in the development of implantable medical devices. A brief introduction on MEMS technology is followed by a premise for the need of MEMS-based implantable devices. The chapter is structured into MEMS sensors (pressure, stress and strain, inertia) and MEMS actuators (drug delivery, electrical stimulators) and their applications in vivo. Recent research efforts in developing implantable MEMS devices are also discussed for each device described. A brief rundown on the challenges faced due to biocompatibility of materials is presented. Lastly, the chapter summarizes the key challenges associated with developing MEMS devices for in vivo applications.

Key words

bio-MEMS

in vivo

drug delivery

biocompatibility

implantable medical devices

11.1 Introduction to MEMS

In 1959, a visionary lecture entitled ‘There is plenty of room at the bottom’ by physicist and Nobel laureate Dr Richard Feynman (Feynman, 1960) motivated the development of a whole new technology that is today popularly called Microsystems Technology (in Europe), Micromachines (in Japan) and microelectromechanical systems (MEMS) in the USA. MEMS technology has essentially evolved from the IC (Integrated Circuits) industry by incorporating mechanical parts into micrometer-scale electronic devices. MEMS is a combination of semiconductor electronic chips and micron-feature mechanical parts such as valves, pumps, mirrors, channels, lenses, heaters, etc. MEMS is a highly interdisciplinary field that includes material science, chemistry, biotechnology, optics and electronics.

A typical MEMS device consists of a sensing or actuating element supported by on-chip electronic circuitry for data processing and in some cases communication. MEMS technology first witnessed successful commercialization in the early eighties in the automobile industry through applications such as accelerometers and pressure sensors in airbags and vehicle control units (Wu et al., 2004). In the last decade, advancements in fabrication technology have opened up promising avenues for MEMS applications in telecommunications (Walker, 2000), data storage (Eleftheriou et al., 2003), biotechnology (Madou and Gurtner, 2002), agriculture (Dell et al., 2009), energy harvesting (Horowitz et al., 2006), seismology (Chen et al., 2005), aerospace (Osiander et al., 2006) and the health care industry (Wise, 2006). In particular, the biotechnology and health care industry has witnessed a phenomenal growth in MEMS devices.

In health care, MEMS devices have been used in vivo as sensors that assist during surgery, long-term sensors for prosthetic devices and remote sensor arrays for real-time collection of physiological data (Eatony and Smith, 1997). MEMS actuators have been successfully used in vivo for drug delivery (Staples et al., 2006). For in vitro applications, a further specialization of MEMS called microfluidics is used for bio-particle identification, gene sequencing, drug discovery in pharmacology and for point-of-care diagnostics such as pathogen detection (Whitesides, 2006). Research in the MEMS field has grown exponentially in the last two decades.

In comparison to their macro-sized counterparts, MEMS devices offer distinct advantages. MEMS devices use reduced energy and materials in production, and integration into a compact unit leads to significant size reduction, which is a critical factor in aeronautical and space exploration applications as well as implantable sensor and actuator systems. Scaling laws establish the increased surface area to volume ratio as the devices are scaled down (Decuzzi and Srolovitz, 2004). This can significantly reduce power requirements and increase selectivity and sensitivity. Smaller size means batch production, which significantly reduces the per-head cost. These and many more advantages have led to extensive use of MEMS technology in most commercial devices such as cellular phones and automobiles.

For medical devices, the advent of MEMS and microfluidic devices promises to change the way health care will be perceived in the near future. Diagnostic devices are now moving from large, expensive analytical instruments in a laboratory to patient-centric, tiny, disposable chips that provide instant results (Whitesides, 2006). The disposable point-of-care devices use small reagent samples of biologically derived fluids such as blood, saliva and urine, avoiding the need for skilled personnel to collect and store bio fluids. The disposable chips also circumvent the problems due to sterility and patient safety. Faster turnaround times and less invasive/painful devices are advantageous due to the size reduction.

Processing technology for MEMS devices has been extensively adopted from IC processing. However, the use of new materials such as polymers has led to the evolution of the processing technology to create new techniques. MEMS sensors and actuators are fabricated using standard microfabrication processes, which include thin film deposition, photolithography and etching. A traditional MEMS process starts with a silicon or glass substrate on which repeated sequences of a combination of bulk micromachining (a subtractive process where the substrate is selectively etched away to form cavities and membranes) and surface micromachining (an additive process where layers of thin films are deposited and patterned to form the desired structure) is used to create the desired microscopic features (Franssila, 2004). Typically, silicon (single-crystal and poly-crystalline silicon) is used as material for MEMS devices due to its excellent mechanical properties coupled with its ability to induce electrical conduction and piezoresistivity using appropriate dopants and doping techniques. Borosilicate glass is also a popular choice for substrate when optical detection is used. Photosensitive polymers and oxide films are used as sacrificial materials to achieve selective deposition and etching. Figure 11.1 presents a schematic of a typical MEMS fabrication process flow. Interested readers are directed to literature sources (Franssila, 2004) that describe the microfabrication processes in greater depth.

11.1 Lithographic patterning process: (a) oxide-film deposition; (b) photoresist application; (c) UV exposure through a photomask; (d) development of resist image; (e) etching of oxide and (f) photoresist removal.

11.2 Requirements for in vivo MEMS

Implantable medical devices have largely been successful as passive prosthetic devices such as orthopaedic implants and dental implants (Webster, 2001), heart valves (Cribier et al.,2002) and vascular replacements (Greislertt et al., 1996). Active medical implants such as cardiac pacemakers and defibrillators (Braun et al., 1999) have also found success, but there still remains a huge potential for monitoring, regulation and stimulation of processes in the various human body organ systems using implantable sensors and actuators.

MEMS devices have more recently been successfully implemented as implantable sensors and actuators. One of the key factors that has influenced the development of implantable MEMS sensors and actuators is the need for online and continuous monitoring of vital statistics of physiological variables in the human body using devices of reduced form factor (Grayson et al., 2004). Conventional diagnostic tools provide an instantaneous assessment, whereas a detailed log of the variations of in vivo conditions would provide a better understanding of the physiological processes and drug interactions. Implanting MEMS devices provides a host of advantages, which include size reduction, low power requirements and patient mobility. The ability of a MEMS sensor to transduce a physical, chemical or biological event into an electrical signal makes it easier to interface to the electronic world. The digital capabilities of MEMS also offer better temporal control over conventional drug delivery systems based on polymeric disintegration and oral consumption. MEMS implantable sensors have found applications in monitoring of blood pressure (Cong et al., 2006), glaucoma (Chen et al., 2008), muscular contraction (Fujita et al., 2011), the vestibular system (Shkel and Zeng, 2006), compressive stresses exerted on the spine (Thaysen et al.. 2002), pH in blood and tissue (Chen et al., 2006), etc. MEMS actuators are being tested for in vivo drug delivery systems (Ibrahim et al., 2007) and microelectrode stimulators (Wang and Wise, 2008) for muscular contractions.

Decades of research have gone into the characterization of the sensing and actuation principles in MEMS devices, and they are well established today. MEMS sensors for in vivo applications require high efficiency and sensitivity, so that power requirements can be reduced leading to longer durability. Integration of these micro devices into a new environment, in this case, the physiological environment, remains the biggest challenge. Biocompatibility of the materials in contact with the surrounding tissue is of high priority. Novel strategies and approaches are required for packaging and materials processing to improve implant life and biocompatibility (Kotzar et al., 2002). For applications requiring extended periods of data collection and monitoring, wireless transmission of information to an external site is required, where data can be recorded and analysed in real-time (Chen et al., 2008). This chapter provides the reader with an overview of the above-mentioned aspects.

11.3 In vivo physiological MEMS sensors

This section introduces the reader to the development of MEMS sensors for in vivo applications. Pressure sensors, which are the most ubiquitous among MEMS sensors, are discussed in depth along with stress, strain and inertial sensors. Various sensing strategies such as capacitance-based, piezoelectric, inertial and potentiometric are discussed. The adaptation of these sensors for in vivo application is presented with examples. A description of the system design, road map to sensor development, wireless communication to external receiver and other challenges faced is presented.

11.3.1 Pressure sensors

Pressure sensors, along with temperature sensors, are among the most ubiquitous sensors and are used extensively on a daily basis. Pressure sensors are used to directly measure pressure, as in a microphone or hydrophone, or can be used to indirectly measure variables such as altitude, speed of fluid/ gas flow, calibration, etc. Many different engineering principles and physics phenomena are employed to sense pressure. Piezoresistive and piezoelectric properties of materials, potentiometric, electrostatic or magnetic change due to displacement of a diaphragm, as well as optical deflection have been used to sense pressure (Eatony and Smith, 1997). Other techniques such as resonance shift, thermal conductivity, thermal imaging and ionization have also been employed for pressure sensing (Eatony and Smith, 1997). While each of these principles offers advantages such as sensing range, sensitivity and remote sensing, choosing the right sensing principle is mostly based on the critical requirement of the application such as area of implantation, expected implant life (durability), integration with wireless communication, power requirements, etc.

MEMS pressure sensors use the same sensing principles and are a miniaturized version of their macroscopic counterpart. Deformable diaphragms are the most commonly used MEMS pressure sensors coupled with capacitive or piezoresistive signal transduction. Here, pressure is determined by the degree of deformation of the diaphragm due to the application of the pressure and the deformation is measured as a change in capacitance or resistance. The obtained pressure value is always referenced to a known reference pressure. Figure 11.2 provides a schematic of a typical deformable membrane pressure sensor (Eatony and Smith, 1997).

11.2 I llustration of a cross-section of a typical pressure sensor diaphragm. Dotted lines represent the undetected diaphragm.

Extensive efforts are underway to apply MEMS pressure sensors as implantable sensors. This section describes examples of MEMS pressure sensors developed for in vivo applications. A surface-micromachined implantable wireless MEMS Intraocular Pressure (IOP) sensor has been developed by Chen et al. (2008) to assist in identifying glaucoma, an optical neuropathy characterized by high ocular fluid pressure in the eye that progressively damages the optic nerve and can result in permanent loss of vision. Parylene was used as a functional material as well as the structural material. Biocompatibility of parylene also provides advantages for integration into the host environment. The first generation device (Meng et al., 2005) consisted of a mechanical pressure sensor that comprised a centrally supported, freestanding parylene spiral tube.

When a pressure differential is generated across the structure, a bending moment is created, forcing an in-plane radial and angular deformation. This effect was monitored by visually tracking the movement of the indicator tip. Deformation that resulted was linearly related to the pressure differential. While this device demonstrated proof of concept, visual detection of the deflection and lack of efficient suturing protocols impeded the actual testing in an implant scenario.

The second generation (Chen et al., 2008) of the device that overcame drawbacks of the predecessor was designed using an electrical LC tank resonant circuit to facilitate passive wireless sensing using an external interrogating coil connected to a readout unit. Two surface-micromachined sensor designs incorporating variable capacitor and variable capacitor/ inductor resonant circuits were used to realize the pressure-sensitive components. The pressure-sensitive capacitor was realized with a flexible diaphragm chamber integrated with parallel metal plates. The electrical LC tank circuit is an inductor-capacitor coupled circuit whose values can be set to a designed resonant frequency. The changes in capacitance due to IOP variations cause a change in the resonant frequency of the LC circuit, which can be characterized by an external reader coil that is inductively coupled to the implanted sensor. This configuration thereby provides continuous wireless monitoring of the IOP. For further details on the complete fabrication process, the reader is directed to Chen et al. (2008).

The implant device dimensions of 4 mm × 1 mm provides a form factor small enough for implantation. The ex vivo bench testing of the system has so far revealed a sensitivity of 7000 ppm/mmHg with a high resolution of 1 mmHg. The use of parylene as the structural material provided biocompatibility, which was verified through a six months animal implant study using a live rabbit eye model. To study biocompatibility, the fully packaged system was implanted into the pars plana site in the eyes of two rabbits. The implants were studied for a period of six months and no inflammatory response or tissue encapsulation was observed. To test sensor performance, the pressure sensor was implanted into an enucleated porcine eye under the cornea. The results for the sensing, although not ideal due to the ex-vivo nature of the testing, have shown promise for better implantation protocols and testing.

MEMS pressure sensors have also been used for in vivo blood pressure monitoring. Cong et al. (2006) have developed an implantable capacitance-based MEMS blood pressure sensing system for real time monitoring of blood pressure. The system uses a biocompatible soft silicone elastic cuff wrapped around a blood vessel coupled with the MEMS capacitive sensor for pressure sensing and low power integrated electronics transmits the signal for real time data collection. The sensor consists of an edge clamped silicon diaphragm over a vacuum cavity that deflects under an applied pressure. The deflection changes the capacitance that can be correlated to the applied pressure. The sensor functioned at a sensitivity of 1 fF/mmHg. Similar to the previously discussed IOP sensor, this system is coupled with a tuneable LC tank circuit to provide for wireless transmission of data to an external recording system. The critical issue of power requirement and, hence, sustainability is well addressed in this work. The integrated electronics for the sensor are designed to convert an incoming radio-frequency (RF) signal into a stable 2 V DC supply to power the entire system. This helps in creating a standalone battery-less implantable MEMS device. While the dimensions of the ASIC (Application Specific Integrated Circuit) architecture are 2 mm × 2 mm, the actual size of the sensor is 0.4 × 0.5 × 0.4 mm3. This form factor provides realistic feasibility for integration with blood vessels. A noticeable drawback, however, is the requirement of the external coil in the proximity of the cuff to power the electronics of the cuff.

Another successful application of implantable MEMS pressure sensors is the monitoring of urinary bladder pressure. This is a critical application as urinary incontinence affects more than half the population of the world above age 60. Cong et al. (2009) have developed a wireless micromanometer system that is designed to measure bladder pressure and telemetrically transmit the data to an external receiver. The MEMS system uses a commercially available MEMS pressure sensor (Silicon Microstructure Inc., SM5102, CA, USA) based on the piezoresistance change of a deformable membrane. The system consists of the MEMS sensor combined with a rechargeable battery, which is implanted under the bladder mucosa layer to chronically monitor the bladder pressure. An ASIC architecture provides the support electronics to transmit the data to an external device. The two external components are the RF receiver and the wireless battery charger. The entire system is designed to fit into a capsule measuring 7 mm wide × 3 mm thick × 17 mm long. While the system design has been demonstrated as proof of concept, it remains to see if the system can adhere to long-term biocompatibility requirements.

The development of various other implantable MEMS pressure sensors (Castro et al., 2007; Chen et al., 2007; Young, 2009) for a variety of in vivo applications is currently underway. With capacitance-based MEMS pressure sensor technology maturing, and achieving successful commercialization in the automotive industry as principle components of safety air bags, the tendency is shifting towards using these commercially available MEMS sensors along with ASICs to create a complete system. The biggest challenge, however, still lies in the efficient packaging of the system for safe implantation into the body. Many of the well-characterized biomaterials that were used for their structural properties are now being investigated for their functional capabilities to achieve easier biocompatibility of the MEMS devices. Commercialization of the systems discussed above and others will require further extensive characterization and approvals from the pertaining governing bodies that maintain highly regulated standards for approval of implantable devices.

11.3.2 Stress and strain sensors

Mechanical loads generate stress and strain of varying intensity on different bones in the human skeletal system. The concentration of stress and strain is a function of the material properties of the bone. Osteoblastic cells, which sense the stress and strain induced on the bones, alter the matrix that forms the bone in cases of excessive stress to alleviate the induced stress. This can result in significant changes in bone properties such as density and porosity and can lead to bone deformations. Monitoring bone structure formation and bone regeneration are therefore important for clinical management of skeletal trauma and early detection of the onset of bone related diseases such as osteoporosis and osteoarthritis. Bone stress monitoring can also help in determining the distribution of stresses for patients with prosthetic implants, which will aid in the design of better prosthetic devices. Current methods are based on imaging techniques such as X-ray absorptiometry and quantitative computed tomography from which bone quality and strength is inferred (Alfaro et al., 2009). Furthermore, the data obtained is not real-time and does not provide much information on temporal variations of stresses exerted on the bones during activities on a daily basis.

Conventionally, metallic strain gauges are used to quantify stress and strain (Hoffmann, 1989). The strain gauge is a device that, when subjected to a physical strain, induces a change in its electrical resistance. The resistance is usually measured using a Wheatstone bridge. The change in resistance is due to the change in the physical geometry of the sensor. True measurement of strain requires a close bond between the strain gauge and the object whose strain is to be measured. An adhesive is best for providing the very close bonding needed between the measurement object and the strain gauge. Apart from metallic strain gauges, mechanical, semiconductor and capacitive strain gauges are used when temperature constraints are likely to offset linearity in metal strain gauges.

In the MEMS domain, strain gauges are realized using several sensing principles such as piezotronic (Zhou et al., 2008), resonance shift tuning forks fabricated from silicon carbide (Azevedo et al., 2007), piezoresistive (Alfaro et al., 2009), electrostatic (Azevedo et al., 2008) and polymer-based (Thaysen et al., 2002). The fabrication of thin films of strain sensitive materials provides good adhesion to the substrate, thereby providing efficient transduction. MEMS-based sensors are good candidates for implantation into the vicinity of bones for continuous monitoring of stress and strain exerted at the micro-scale. Stresses developed at the micro-scale due to bone-implant stress, bone-tissue stress, friction and micro cracks when detected early can be used as a prognostic tool to avoid development of irreparable damage. This is especially critical for older people for whom bone regeneration is significantly hampered.

Alfaro et al. (2009) have reported the development of MEMS-based multi-axis sensors for direct measurement of bone stress at the micro- scale (Fig. 11.3). The multi-axis sensing of stress provides information to objectively evaluate the healing of bone, given its anisotropy and complex microstructure. The sensor system uses an array of piezoresistive sensors that are in contact with the bone surface. Each sensor in the array is connected in a Wheatstone bridge configuration and the corresponding signal from each sensor is routed through an interconnect stack to data collection. The MEMS sensor is supported by CMOS circuitry for data collection and transmission. The entire assembly is housed in a 3 mm × 3 mm × 0.3 mm packaging providing a reasonable form factor for implantation. The device was designed for a maximum load of 200 kPa while maintaining a high sensitivity of 100 Pa.

11.3 (a) Envisioned implantable CMOS-MEMS multi-axis stress sensor. (b) Visualization of a 60 μm × 60 μm × 60 μm tall silicon post indicating the location of piezoresistors in the silicon under the oxide layer. The interconnecting beam consists of metal and oxide layers. (c) SEM of the bone stress sensor after micromachining the 0.35 μm BiCMOS chip.

Glos et al. (2010) have developed a MEMS compressive stress sensor for evaluating the pressure exerted on the annulus of the intervertebral disc. The study of compressive stress in the annulus of the intervertebral disc is essential in the design of advanced surgical techniques that require biomechanical gradients such as spine growth modification. The system consists of a commercially available, capacitance-based MEMS pressure sensor (Silicon Microstructure Inc., CA, USA), which is modified and packaged to protect the electrical system from the biochemical and biomechanical environment rendering it biocompatible. Figure 11.4 presents a picture of the implantable device housing the sensor dies in a thin, L shaped, metallic carrier. The metallic carrier was made of stainless steel. The commercially available MEMS pressure sensor was unpackaged and bonded to the steel carrier using a suitable adhesive and electrical leads were created using a flexible ribbon cable. The entire assembly was packaged by coating with Parylene-C, a biocompatible polymer, using a chemical vapour-deposition technique. In vivo tests were conducted by implanting the device into a porcine animal model and found to be largely successful in maintaining signal output and sensitivity (Fig. 11.4). Stress and output voltage from the pressure sensor were linearly correlated over a range of 0–1.8 MPa with less than 5% change in sensitivity. The form factor (0.9 mm thick, 3.2 mm wide and 30 mm long) is small enough to be accommodated into the vertebral disc space. Further implant studies towards characterization of longerterm stability and host response will yield more information and thus better design.

11.4 (a) Silicon pressure die (SM5112, Silicon Microstructures Inc., Milpitas, CA) incorporated into custom stainless carrier including hermetic barrier coating (Parylene-C, vapor deposited, conformal). Excised porcine thoracic spine segment: (b) Anterior view showing locations in left and right sides of intervertebral disc at the mid-transverse plane. Sensors attached to vertebra with suture anchors 3–5 mm from one end plate; sutures looped twice around sensor leg and secured with surgical knot (denoted by arrows). (c) Transverse section showing placement in lateral aspects of annulus. Sensing element (denoted by dotted circle) in mid-coronal plane of the disc, orthogonal to longitudinal axis of spine.

11.3.3 Inertial sensors (accelerometers and gyroscopes)

Accelerometers are instruments used to measure rate of change of acceleration. MEMS-based accelerometers were one of the earliest MEMS products to find successful commercialization (Wu et al., 2004). MEMS accelerometers are the critical sensing components in automotive crash sensors that trigger air bags. The physical mechanism underlying MEMS accelerometers may be capacitive, piezoresistive, piezoelectric, optical, electromagnetic or one of many others. Capacitance-based accelerometers are, however, the most successful and popular mechanism in the MEMS domain due to the relatively simple design and passive operation.

Capacitance-based MEMS devices have several attractive features. Capacitors can be used as both sensors and actuators while maintaining excellent sensitivity and feedback control. The sensing, which is based on change in capacitance is insensitive to temperature variations (Eatony and Smith, 1997). Capacitive sensing is independent of the base material and relies on the variation of capacitance when the geometry of a capacitor is changing. The basic equation for obtaining the capacitance is given as:

[11.1]

where, εo and ε are the permittivity of free space and the permittivity of the material between the two parallel plates, respectively, A is the cross sectional area and d is the distance between the two parallel plates. Acceleration sensing is essentially based on changes in either d or A. A typical MEMS accelerometer uses a movable proof mass with plates, attached through a mechanical suspension system to a reference frame, as shown in Fig. 11.5 (Krishnamoorthy et al., 2008).

11.5 MEMS accelerometer optical near-field resonant displacement sensor based on vertically stacked sub-wavelength nano-gratings. The nano-gratings are attached to electrostatic actuators to control their motion and characterize their displacement sensitivity.

MEMS accelerometers are now finding applications in personal electronics such as mobile phones, laptops and gaming devices. MEMS accelerometers have also found extensive use in seismology to quantify the intensity of earthquakes (Chen et al., 2005).

Gyroscopes are devices that measure angular rate, and this information is often then processed to provide a measure of orientation. In a closed loop system, this information can be used in a feedback loop to maintain desired orientation, for example, in a stabilisation platform (Schiff, 1960). Gyroscopes work on the principle of conservation of angular momentum. The first gyroscopes were mechanical spinning wheels that had an axle that could freely change orientation. Gyroscopes based on other operating principles exist such as vibrating rate gyroscopes, solid-state laser gyroscopes, fibre optic gyroscopes and quantum gyroscopes.

Gyroscopes find application in everyday activities and are critical components of many devices. Gyroscopes are a critical component of navigation assisting equipment in airplanes and ships where they are used in stabilizers to maintain balance during motion (Dean et al., 2005). Gyroscopes are also used in cars for stabilisation control, anti-lock braking systems and activation of roll cages (Sassen et al., 2000). The famous Hubble Space Telescope also uses three gyroscopes to steadily change orientation to view celestial bodies. In recent times, gyroscopes have found extensive application in mobile phones and virtual reality devices for gaming that has broadened their functionality beyond expectations (Lane et al., 2010).

MEMS gyroscopes are electronically packaged microchips that utilize the resonance property of a vibrating structure. In order to maximise the Coriolis forces, structures can be induced to vibrate, and the motion due to Coriolis forces can be measured and related to the angular rotation. MEMS gyroscopes can be constructed as two-dimensional (2D) structures like tuning forks, spinning wheels, vibratory structures, or axis-symmetric structures like rings or discs. There is a range of functional materials used, and the structures have been designed so that they can be fabricated using standard MEMS micromachining techniques (Xie and Fedder, 2003). Sometimes, MEMS gyroscopes are integrated with MEMS accelerometers to obtain data for all six degrees of freedom.

In the human body, the anatomical system that is analogous to the gyroscope and accelerometer is the vestibular organ in the inner ear, which is responsible for providing the brain with information on the motion and orientation of the body (Weinberg et al., 2006). The vestibular system consists of three semi-circular canals, the utricle and the saccule. The semi-circular canals are responsible for rotational head movements, while the saccule and utricle are responsible for linear movements. Sensory hair cells present in the three inner ear organs send impulses through the vestibular nerve bundle to the brain, where information about head movement is combined and interpreted. The vestibular system of the inner ear provides cues about self-motion that help stabilize vision during movement. These cues also enable us to orient ourselves with respect to our surroundings, which helps us to stand and walk.

A loss of balance due to failure of the vestibular system in the human body can be a serious concern. This is seen predominantly in older people, who are at risk of falling; recovery from a fall is also more difficult in older people due to age-related diminished healing. Vestibular system failure is also seen in patients with severe injuries and trauma to the head and neck.

For implantable medical devices, MEMS gyroscopes and accelerometers can be used as a vestibular prosthetic device to replace a damaged vestibular organ. Several research groups are working on developing implantable MEMS gyroscopes and accelerometers. Weinberg et al. (2006) have envisioned a fully implantable MEMS device, which uses three gyroscopes and three accelerometers providing real time data of the six degrees of motion. The instrument sensor assembly includes readout electronics, digitizers and a bubble level mechanism for aligning the instruments with the patient’s natural or comfortable vertical. The three commercially available accelerometers (Analog Devices, MA, USA) had good stability in terms of thermal sensitivity (3 milligravity (mg) per°C) and noise (0.23 m g√Hz). The three gyroscopes from Silicon Sensing Systems (Plymouth, UK) had a thermal sensitivity of 650 deg/h/°C and noise at 470 deg/h√Hz. Algorithms to identify tilt and convert the data into appropriate electrical impulses have been successfully developed. The wearable vestibular prosthesis has shown promise as both a laboratory-testing tool and ultimately as a rehabilitation prosthesis. Thus far, the most significant results were obtained in standard clinical tests where balance-impaired subjects were deprived of vision and proprioceptive inputs. In single-axis tests, balance impaired subjects who fell when not aided were able to stand with the prosthesis.

Shkel and Zeng (2006) have developed a prototype MEMS cochlear and vestibular prosthetic device that can sense motion with precision and deliver signals to the central neural system thus mimicking the dynamic vestibular function. The device includes three main functional units – a sensing unit, a pulse generator and a stimulator. The sensing unit uses a single-axis MEMS gyroscope that has a form factor of 2 mm × 2 mm. The power requirement to drive the gyroscope is about 10 mW, which is comparable to commercially available gyroscopes (Murata and Analog Devices Inc.). However, better circuit design and power management protocols should help realize power consumption in the μW range, making it feasible for long-term implantation and usage. An oscillator circuit drives the proof mass, which is the active component of the sensor. When subject to rotation with angular velocity, the proof mass will be subject to the Coriolis force. The resultant Coriolis force is perpendicular to both the input rate and the instantaneous radial velocity in the drive direction. This force produces a motion of the proof mass in a direction perpendicular to its initial oscillation. The deflection is proportional to the angular velocity. Voltage signals are generated proportional to the angular acceleration of the head. Appropriate algorithms process the input voltage signal and generate current pulses that stimulate the vestibular neurons. The proposed implantable MEMS vestibular prosthetic device, although still in the initial implementation stage, has shown promising preliminary results and is currently being investigated on animal models (Zurcher et al., 2007).

11.4 In vivo MEMS actuators

This section introduces the reader to MEMS actuators used for in vivo applications. While MEMS sensors predominantly have passive components, MEMS actuators involve active components with mechanical parts. MEMS actuators are used as micro-valves (Chakraborty et al., 2000), micro-pumps (Nguyen et al., 2002) and micro-grippers (Vasudev and Zhe, 2008) by employing various actuation mechanisms such as electrostatic, piezoelectric, piezoresistive, shape memory, pneumatic and thermal mechanisms (Staples et al., 2006). For in vivo applications, the biggest beneficiaries of MEMS actuators have been in the electrical stimulator (Wang and Wise, 2008) and drug delivery domains (Elman et al., 2008; Prescott et al., 2006; Rao et al., 2005).

11.4.1 Drug delivery systems

Drug delivery is the process of delivering a pharmaceutical or therapeutic agent into a living organism. The method of delivery may be the traditional oral form, dermal, nasal, vaginal, ocular or rectal. Certain therapeutics and vaccines that are gene based are delivered via injections or infusions to avoid degradation due to enzymatic action. A more target-specific delivery of a drug is sought when the therapeutic agent, as in the case of chemotherapy, is to be administered only to the area of interest (Arap et al., 1998). Another factor to consider during drug delivery is that of controlled release versus conventional release. Controlled release is the delivery of the drug at a controlled rate for an extended period (Gupta et al., 2002). There are several advantages to controlled release such as maintaining a constant concentration of the drug during critical conditions as compared to conventional release where the drug is periodically administered. In conventional drug release, the concentration of the drug varies by a large margin, with higher than required concentrations at the time of administering, tapering down as time progresses. Hence, extreme care has to be taken to avoid overdose at the time of administering. So far, drug polymer conjugates and biodegradable microspheres have been used for controlled and sustained drug release (Jeong et al., 1997). While these methods have found considerable success, they may still be classified as passive delivery vehicles since they rely on time-dependent degradation chemistry of the polymer materials in order to release the drug. A more active delivery mechanism can be implemented using MEMS technology to implant the device and then telemetrically engage the device on demand. MEMS-based drug delivery devices are mostly a combination of micro-pumps/actuators, micro-valves and micro-reservoirs (Ibrahim et al., 2007). The micro-pumps utilize actuation principles such as electrostatics, piezoelectric, pneumatic, thermo-pneumatic, magnetic and shape memory alloys. Micro-valves are used to control and regulate fluid motion during delivery. The micro-reservoirs are typically an array of micromachined reservoirs containing a single drug or a combination of drugs. The drug-loaded reservoirs can be individually addressed to release the drug, thereby allowing for different combinations of drug mixtures to be delivered.

Elman et al. (2009) have developed a fully implantable MEMS drug delivery device for rapid delivery in cases of ambulatory emergency. The MEMS device can be used for subcutaneous, intraperitoneal, intramuscular or trans-dermal delivery. The device architecture is a 3-module system consisting of a single reservoir containing the drug, a hermitically sealed membrane and a layer of micro-resistors for actuation. Vasopressin, an antidiuretic hormone (ADH), which plays a key role in the regulation of water, glucose and salts in the blood, was used as a model to demonstrate the device working in vitro. The actuation layer is defined by microresistors, which, once activated, rapidly and locally heat the contained fluid to generate bubbles (Fig. 11.6a). The increase in pressure caused by the bubbles ruptures the membrane and propels the contained solution out of the device, allowing the delivery of the drug. The ruptured the membrane allows the drug to be dispensed at a rate of 20 μL in 45 s. The construction materials of this device (silicon, silicon nitride, silicon dioxide, gold and titanium) create a completely biocompatible system. The power requirement for the electro thermal actuation was 25 W/mm2 allowing for sustainable usage after implantation. The complete system was packaged to a form factor suitable for implantation (Fig. 11.6b).

11.6 (a) Cross-sectional render of a MEMS drug delivery device showing the three layers: membrane layer (A), reservoir layer (B), actuation layer (C). (b) Final packaged device. (c) Side view of device showing release of methylene blue into a solution upon activation.

MicroCHIPS Inc. (Bedford, MA, USA) (Prescott et al., 2006) has developed a programmable polypeptide delivery system using an implantable multi-reservoir microchip device. The micromachined device consists of 100 individually addressable micro-reservoirs (Fig. 11.7). Upon implantation, each reservoir of 300 nL volume can be addressed individually and remotely. The addressability of the individual reservoirs provides a very good tool for researchers in the field of drug discovery to try combinational drugs, with different drugs in different reservoirs. The micromachined reservoirs are opened using an electro-thermal actuation mechanism where a thin metal membrane that covers each reservoir is instantly removed by localized resistive heating from an applied current. The electro-thermal approach has also been implemented previously by Maloney et al. (2005) for implantable drug delivery devices. Each reservoir was filled with 25 mg of lyophilized leuprolide in a matrix of solid polyethylene glycol. A current of about 0.3 A was used to open the reservoir. The response of the device is almost instantaneous (5 μs). Each microchip measures 15 mm × 15 mm × 1 mm. The device has been successfully tested in vivo for over six months implantated in a bovine animal model.

11.7 I mages of the microchip reservoirs and implantable drug delivery system. (a) Front and back of the 100-reservoir microchip. (b) Representation of a single reservoir. (c) Electronic components on the printed circuit board (PCB) in the device package. (d) The assembled implantable device.

Similarly, Rao et al. (2005) have developed an implantable MEMS drug delivery system for chemotherapy by integrating a subcutaneous reservoir, an in-plane silicon pump and supporting electronic circuitry for local and centralized delivery of therapeutic agents. The silicon micropump functions through the use of a stroke amplifier, movable diaphragm and a valve. Localized heating of the diaphragm and lever using a current source amplifies the stroke of the pump. A deflection of 30 μm at an applied voltage of 12 V resulted in a pumping volume of 3.6–12.6 nL of the drug and compression ratio of 3.5. While the system has been integrated and the functionality has been demonstrated, further testing is directed at using the system in vivo.

MEMS-based drug delivery systems have demonstrated great potential in a vast number of applications that require controlled and accurate delivery of therapeutic agents without the need for direct medical intervention. While the devices have been demonstrated as proof of concept and in preliminary implant studies, further investigation is required to address issues such as re-usability of the implanted device, refilling of the drug for long-term usage and control of degradation of the therapeutic agent in the micro-reservoirs. Actuation methods such as electro-thermal and electro-chemical are currently the popular choice. The electro-thermal method (Maloney et al., 2005) has the advantage of being independent of the chemistry of the environment surrounding the device and is many times faster than the electro-chemical method. In any electrically-based actuator, keeping the power requirements low is always a matter of critical importance.

11.4.2 Electrical stimulators

An in depth understanding of the dynamics of the neuromuscular system has led to the development of therapies for muscular dysfunction using electrical stimulation (Leuthardt et al., 2006). Electrical stimulation that mimics the action potentials sent from the brain to innervate muscle is the basis for electrical stimulation based therapy. An example of successful implementation is the electrical stimulation of the heart muscle using pacemakers and defibrillators (Braun et al., 1999). The application of electrical stimulation has great potential to be applied to other parts of the body such as the spinal cord, brain, auditory system and any organ that has lost function due to nerve damage. Electrical stimulation can provide much needed relief in cases where administration of drugs has no effect, where trauma has cut off neural contact, etc.

Electrical stimulators are typically a bundle of wires that taper down in diameter at the end. This large size significantly hinders access and implantation for long periods. The relatively large size also limits the density at which electrical stimulation is achievable. Electrical stimulation of the brain typically requires stimulation of a specific part of the brain in an isolated manner. These constraints have encouraged the development of MEMS-based electrical stimulators that can be implanted and used for high-density stimulation (Wu and Bernstein, 2006). Microelectrode arrays can be fabricated using standard microfabrication techniques and can be individually addressed on demand using on-chip electronics. This section presents a few examples of successful electrical stimulators developed using MEMS technology.

For patients with profound hearing loss, Wang and Wise (2008) have developed a MEMS-based cochlear prosthetic device that uses a micromachined hybrid electrode array integrated with a position sensor to restore hearing capabilities. The fully implantable system has an array of single-addressable electrodes, which, thereby, provides high-density stimulation. The electrode array is integrated with position sensors that can be used for guidance during implantation. The guidance from the position sensors is critical since the tympanic canal in the cochlea is narrow with the canal tapering down from 1 mm to 200 μm along a 35 mm long channel. If damaged (during implantation), hair cells, which are the sensory receptors of the auditory system, are not replenished, hence extreme caution and guidance are required. The position sensors are piezoresistive strain gauges realized from microfabricated thin film of high-impedance poly-silicon implanted with arsenic. The form factor of the electrode array is 8 mm long × 2 μm thick, and the width tapers from 600 to 200 μm thereby allowing implantation of the electrodes into the inner depths of the tympanic canal. The signal processing chip (2.4 mm × 2.4 mm), which drives the electrode array, is connected using a thin polymeric cable. The entire system is operated at about 2.5 V, keeping power requirements in an operable range.

Similarly to electrical stimulation of the vestibular nerve bundles, electrical stimulation of the neural network or neurons in the brain is used to study the electrical activity of the brain. Also called Deep Brain Stimulators (DBS) (Wu and Bernstein, 2006), they are used as a therapeutic technique to activate neurons that have been damaged due to injury or in the case of diseases such as Parkinson’s disease. The electrical stimulus is delivered to the neurons using micro-actuators integrated with microprobe tips.

Researchers at IMEC, Belgium (Musa et al., 2009) have developed an implantable micromachined micro-actuator that can be integrated with micro-needles and microprobe tips to make micrometer and nanometer manipulations of in vivo brain electrodes (Fig. 11.8). The microelectrode array was fabricated on a silicon substrate using a combination of bulk and surface micromachining. An assortment of electrode designs were fabricated and tested to characterize the response for both recording and stimulation. The silicon electrode array had a shaft length of 2 mm and 200 × 200 μm2 cross-section. The micro-manipulators, which had a stroke length of 50 μm and a sizeable force of 195 μN, were used for precision insertion of the probes into the brain at a constant speed of 10 μm/s. The electrical stimulus was designed to provide a 600 μA, 0.2 ms pulse stimulus. In vivo recording and stimulation experiments were performed in the cortex (1 mm × 2 mm) of an anesthetized rat. The functioning of the system was verified with action potentials recorded from the motor cortex of the hind limb area, and electrical stimulus of the same motor cortex resulted in the contraction of the hind limb.

11.8 Optical micrographs of probe tips with different electrode sizes and configurations: tip with electrodes of 50 pm diameter (a), of 25 pm diameter (b), and with C-shaped electrodes (c). Picture of the whole probe (d) and the packaged probe (e).

In conclusion, the neurological system, which is the most complex system in the body, can be better studied using electrical stimulation and recording systems that have a high degree of addressability. The successful implementation of MEMS-based electrical stimulators will encourage the development of similar treatment for other neurological conditions such as epilepsy and obsessive-compulsive disorder.

11.5 Biocompatibility

Currently, MEMS-based devices are being developed for a wide variety of implantable applications. While the functionality of most of these devices is being successfully demonstrated, the biggest challenge remains in the successful integration of the device into the physiological environment (Kotzar et al., 2002). Almost all of the MEMS devices designed for in vivo application are based on the interaction of a functional material with the physiological environment. Therefore, to achieve the desired performance from the device, it is imperative to understand the host tissue’s response to the device and the materials at the material-tissue interface. In an ideal scenario, all the materials used in a MEMS device would be biocompatible. However, in practice, some of the functional materials that are critical to device performance may not meet the required specifications of biocompatibility such as sterility, cytotoxicity, immune response and inflammatory response. We will discuss here the importance of these critical requirements.

In a generic sense, a biocompatible material can be defined as any material that will be readily accepted by the physiological environment with minimum inflammation or toxic response. The human body is designed to identify and eliminate any foreign material that enters the system. The first indication of the presence of a foreign material is the inflammatory response (Black et al., 2006). The inflammation is a non-specific response of the human body to several factors such as tissue damage from trauma, infection, local cell death and intrusion of foreign material. The clinical signs of the inflammatory response include redness, swelling, pain and localized heating of the tissue surrounding the foreign material. These indicators trigger the immune response, which includes the recruitment of phagocytes, natural killer cells, and antimicrobial chemicals to the site of the foreign body to disintegrate the material and initiate tissue repair. In the case that the body is unable to disintegrate the foreign material, there is a tendency to encapsulate the material with proteins and thereby render it harmless. While this response from the body is helpful in integrating biomaterials for bone replacement and mechanical heart valve applications, it can cause serious performance issues in MEMS sensors since the sensing surface needs to be exposed to the physiological environment at all times. Strategies adopted to mitigate these problems involve the consumption of anti-inflammatory and immune-suppressant drugs immediately after implantation. In some cases, patients with implants continue to take immune-suppressant drugs for as long as the implant is present in the body. Implementation of this strategy, however, comes with its drawbacks. A patient on immune-suppressant drugs has a weakened immunity system, making him/her susceptible to attacks from viruses and bacteria. Extreme precaution has to be taken by the patient to avoid such attacks.

Toxicity is another biocompatibility issue that needs to be considered while designing implantable devices (Ratner et al., 1996). While inflammatory and immune response is the host reaction, toxicity deals with the reaction of the material to the host. The dynamic host environment where pH, temperature and tissue constituents vary can induce a reaction from the material such as degradation at the surface. The degradation products, if toxic, can lead to cell death and cause disease (Black et al., 2006). Parameters such as LD50 (Lethal Dose) and LC50 (Lethal Concentration) are used to test toxicity of biomaterials. Statistically, LD)50 is defined as the smallest dose of a substance that can cause the death of 50% of the test population. LC50 is the smallest concentration of a substance that leads to death of 50% of the test population. Degradation may also be a time-dependant factor, hence, affecting implant life. Sterilization of the device prior to implantation is mandatory in order to rule out the introduction of any bacteria, virus or other toxic contaminants into the body (Athanasiou et al., 1996). The sterilization technique is chosen depending on the material. Most MEMS-based devices used for in vivo application have a polymeric encapsulation, for which radiation (UV, ultrasonic) based sterilization is employed. Biofouling (Tao et al., 2008), which is the coating of the material through protein adsorption, can be catastrophic for sensors. Methods to prevent biofouling such as coating the material with anti-adsorption materials, tissue and cell culture over the device for better integration, surface texturing, etc., can be used.

Packaging of the MEMS device will play a key role in isolating the non-biocompatible materials to ensure no physiological contact with incompatible materials. Typically, MEMS devices are packaged with biocompatible polymers such as polyimide, Parylene-C, Teflon®, PDMS, polystyrene, etc. Extensive studies are being conducted into determining the biocompatibility of emerging and existing materials used in MEMS devices (Najafi, 2007).

Popular MEMS materials such as single-crystal silicon, poly-crystalline silicon, silicon oxide (SiO2), silicon nitride (Si3N4), silicon carbide (SiC), gold, titanium and polymers have been tested for biocompatibility (Najafi, 2007). While most of these materials appear to be biocompatible, they cannot be deemed so due to the complex dynamics of the physiological environment; also variations in material processing can affect the response of the material in the physiological environment.

11.6 Conclusions

This chapter has discussed implantable MEMS devices under the broad classification of sensors and actuators. MEMS sensors for pressure sensing, stress and strain sensing, and inertial sensing were introduced. These sensors were realized using sensing strategies based on electrostatic (capacitance), piezoelectric, inertial and potentiometric principles. Implantable MEMS sensors for a variety of in vivo applications such as blood pressure monitoring, urinary bladder pressure monitoring, IOP monitoring, stress monitoring on the intervertebral disks, bone stress measurement, vestibular and cochlear prosthesis were discussed. MEMS actuators for in vivo application such as micro-pumps, micro-manipulators and electrical stimulators were also introduced. The mechanisms used for actuation included piezoelectric, electro-thermal, electro-chemical and small magnitude electric currents. The actuation principle along with application examples of recently developed MEMS systems such as drug delivery systems and micro electrode based high-density electrical stimulators were elucidated in this chapter. Challenges related to applying the developed systems in vivo such as biocompatibility, inflammatory and immune response, toxicity, sterility and the various techniques employed to tackle these problems were discussed in the biocompatibility section.

MEMS technology has witnessed a phenomenal growth over the last decade and has met with significant success in terms of commercialization. While MEMS has found applications in almost all sectors of technology, health care has been the biggest beneficiary. With tremendous potential for growth, MEMS and microfluidics are in the process of significantly changing the way diagnostics and therapeutics will be perceived in the near future. The advantages of micron sized features render many advantages for medical applications. Applied to the field of implantable medical devices, the small size, along with low power consumption and precision, make MEMS devices ideal candidates for in vivo applications.

11.7 Future trends

This chapter has discussed a wide variety of sensors and actuators as applicable to implantable medical devices. Major advances have been made to the existing technology. While the concepts and prototypes have been demonstrated, it is notable that most of these sensors and actuators are still far from being commercialized. A successfully demonstrated concept will have to go through rigorous approvals from governing bodies that regulate the approvals for implantable medical devices. To make a successful transition from laboratory settings to clinical settings further advancement in technology is required to improve the robustness, patient safety, packaging and material processing.

The biggest challenge, however, lies in the successful implementation and sustenance of the implantable devices. Biocompatibility of the materials and packaging will be a critical factor. Direction for future work should be in the study of device performance in vivo over longer periods of implantation.

11.8 Sources of further information

Development of implantable medical devices is an emerging field. With the influx of MEMS technology into implantable devices, a niche area of work has been identified. Since most of the research work is currently being done in research labs, a good source for further information would be peer-reviewed journal articles. Review articles on the state of the art in MEMS-based implantable medical devices are available and have been referenced in this chapter appropriately. Further information on MEMS technology can be obtained from several books such as ‘An Introduction to Microelectromechanical Systems Engineering’, by N. Maluf; ‘Microsystem Design’, by S. Senturia; ‘Micromachined Transducers Sourcebook’, by G. Kovacs. For more information on microfabrication of MEMS devices, the reader is directed to books such as ‘Fundamentals of Microfabrication’, by M. Madou and ‘Introduction to Microfabrication’, by S. Franssila. Key Journals to refer in this field are Journal of Microelectromechanical Systems, Journal of Micromechanics and Microengineering, Biomaterials, Biomedical Microdevices, Journal of Biomechanics, etc.

11.9 References

Alfaro, F., Weiss, L., Campbell, P., Miller, M., Fedder, G.K. Design of a Multi-axis Implantable MEMS Sensor for Intraosseous Bone Stress Monitoring. J. Micromech. Microeng.. 19, 2009. [085016].

Arap, W., Pasqualini, R., Ruoslahti, E. Cancer Treatment by Targeted Drug Delivery to Tumor Vasculature in a Mouse Model. Science. 1998; 279(5349):377–380.

Athanasiou, K.A., Niederauer, G.G., Agrawal, C.M. Sterilization, toxicity, biocompatibility and clinical applications of polylactic acid/polyglycolic acid copolymers. Biomaterials. 1996; 17:93–102.

Azevedo, R.G., Jones, D.G., Jog, A.V., Jamshidi, B., Myers, D.R., Chen, L., X-a, Fu., Mehregany, M., Wijesundara, M.B.J., Pisano, A.P. A SiC MEMS Resonant Strain Sensor for Harsh Environment Applications. IEEE Sens. J.. 2007; 7(4):568–576.

Azevedo, R.G., Huang, W., O’Reilly, O.M., Pisano, A.P. Dual-mode Temperature Compensation for a Comb-driven MEMS Resonant Strain Gauge. Sensor Actuator A. 2008; 144:374–380.

Black, J. Biological Performance of Materials: Fundamentals of Biocompatability, 4th Edition. Taylor & Francis Group; 2006.

Braun, T.C., Hagen, N.A., Hatfield, R.E., Wyse, D.G. Cardiac Pacemakers and Implantable Defibrillators in Terminal Care. J Pain Symptom Manage. 1999; 18:126–131.

Castro, P.F., Conceptión, R., Bourge, R.C., Martínez, A., Alcaino, M., Deck, C., Ferrada, M., Alfaro, M., Perrone, S. A Wireless Pressure Sensor for Monitoring Pulmonary Artery Pressure in Advanced Heart Failure: Initial Experience. J. Heart Lung Transplant.. 2007; 26(1):85–88.

Chakraborty, I., Tang, W.C., Bame, D.P., Tang, T.K. MEMS Micro-valve for Space Applications. Sensor. Actuator.. 2000; 83:188–193.

Chen, M., Glaser, S.D., Oberheim, T., Terra-Scope – A MEMS-Based Vertical Seismic Array. Proceedings, SPIE Symposium on Smart Structures & Materials/ NDE. 2005. [San Diego].

Chen, P.-J., Rodger, D.C., Saati, S., Humayun, M.S., Tai, Y.-C. Microfabricated Implantable Parylene-Based Wireless Passive Intraocular Pressure Sensors. J. Microelectromech. S.. 2008; 17(6):1342–1351.

Chen, Y., Wang, X., Erramilli, S., Mohanty, P. Silicon-based Nanoelectronic Field-effect pH Sensor with Local Gate Control. Appl. Phys. Lett. 89, 2006. [223512].

Cong, P., Suster, M.A., Chaimanonart, N., Young, D.J., Wireless Power Recharging for Implantable Bladder Pressure Sensor. Christchurch. 2009:1670–1673. [IEEE Sensors].

Cong, P., Ko, W.H., Young, D.J., Wireless Less-Invasive Blood Pressure Sensing Microsystem for Small Laboratory Animal In Vivo Real-Time Monitoring. Actuators and Microsystems. 13’ International Conference on Solid-State Sensors. 2006:2002–2005.

Cribier, A., Eltchaninoff, H., Bash, A., Borenstein, N., Tron, C., Bauer, F., Derumeaux, G., Anselme, F., Laborde, F., Leon, M.B. Percutaneous Transcatheter Implantation of an Aortic Valve Prosthesis for Calcific Aortic Stenosis First Human Case Description. Circulation. 2001; 106:3006.

Dean, R., Flowers, G., Sanders, N., Horvath, R., Kranz, M., Whitley, M. Vibration Isolation of MEMS Sensors for Aerospace Applications. J. Microelectronics and Electronic Packaging. 2005; 2(4):223–231.

Decuzzi, P., Srolovitz, D.J. Scaling Laws for Opening Partially Adhered Contacts in MEMS. J. Microelectromech. S.. 2004; 13(2):377–385.

Dell, J.M., Milne, J.S., Antoszewski, J., Keating, A.J., Schuler, L.P., Faraone, L., MEMS-based Fabry-Perot Microspectrometers for Agriculture. Proc. SPIE. 2009. [7319].

Eatony, W.P., Smith, J.H. Micromachined Pressure Sensors: Review and Recent Developments. Smart Mater. Struct.. 1997; 6:530–539.

Eleftheriou, E., Antonakopoulos, T., Binnig, G.K., Cherubini, G., Despont, M., Dholakia, A., Dürig, U., Lantz, M.A., Pozidis, H., Rothuizen, H.E., Vettiger, P. Millipede – A MEMS-Based Scanning-Probe Data-Storage System. IEEE Trans. Magnetics. 2003; 39(2):938–945.

Elman, N.M., Ho Duc, H.L., Cima, M.J. An Implantable MEMS Drug Delivery Device for Rapid Delivery in Ambulatory Emergency Care. Biomed. Microdevices. 2009; 11:625–631.

Feynman, R.P. There Is Plenty of Room at the Bottom. Eng. Sci.. 1960; 23(5):22–36.

Franssila, S. Introduction to Microfabrication. New York: Wiley; 2004.

Fujita, H., Dau, V.T., Shimizu, K., Hatsuda, R., Sugiyama, S., Nagamori, E. Designing of a Si-MEMS Device with an Integrated Skeletal Muscle Cell-based Bio-actuator. Biomed. Microdevices. 2011; 13:123–129.

Glos, D.L., Sauser, F.E., Papautsky, I., Bylski-Austrow, D.I. Implantable MEMS Compressive Stress Sensors: Design, Fabrication and Calibration with Application to the Disc Annulus. J. Biomech.. 2010; 43:2244–2248.

Grayson, A.C.R., Shawgo, R.S., Johnson, A.M., Flynn, N.T., LI, Y., Cima, M.J., Langer, R. A BioMEMS Review: MEMS Technology for Physiologically Integrated Devices. Proc. IEEE. 92(1), 2004.

Greislertt, H.P., Gosselint, C., Rent, D., Kangt, S.S., Kim, D.U. Biointeractive Polymers and Tissue Engineered Blood Vessels. Biomaterials. 1996; 17:329–336.

Gupta, P., Vermani, K., Garg, S. Hydrogels:From Controlled Release to pH Responsive Drug Delivery. DDT. Vol. 7, 2002. [No. 10 May 2002].

Hoffmann, K. An Introduction to Measurements Using Strain Gages. Hottinger Baldwin Messtechnik GmbH; 1989.

Horowitz, S.B., Sheplak, M., Cattafesta, L.N., Nishida, T. A MEMS Acoustic Energy Harvester. J. Micromech. Microeng.. 2006; 16:S174–S181.

Ibrahim, F., Abu Osman, N.A., Usman, J., Kadri, N.A. A Review of MEMS Drug Delivery in Medical Application. IFMBE Proceedings. 2007; Vol. 15:312–315. [Biomed 06].

Jeong, B., Bae, Y.H., Lee, D.S., Kim, S.W. Biodegradable Block Copolymers as Injectable Drug-delivery Systems. Nature. 1997; 388:28.

Kotzar, G., Mark Freas, M., Abel, P., Fleischman, A., Roy, S., Zorman, C., Moran, J.M., Melzak, J. Evaluation of MEMS Materials of Construction for Implantable Medical Devices. Biomaterials. 2002; 23:2737–2750.

Krishnamoorthy, U., Olsson, R.H., III., Bogart, G.R., Baker, M.S., Carr, D.W., Swiler, T.P., Clews, P.J., In-plane MEMS-based Nano-g Accelerometer with Sub-wavelength Optical Resonant Sensor. Sensor. Actuator. A. 2008:145–146. [283-290].

Lane, N.D., Miluzzo, E., Lu, H., Peebles, D., Choudhury, T., Campbell, A.T., A Survey of Mobile Phone SensingAd Hoc and Sensor Networks. IEEE Comm, 2010.

Leuthardt, E.C., Schalk, G., Moran, D., Ojemann, J.G. The Emerging World of Motor Neuroprosthetics: A Neurosurgical Perspective. Neurosurgery. 2006; 59(1):1–14.

Madou, M., Gurtner, C., Bio-MEMS – The Impact of MEMS on Biotechnology in the 21st Century. Device Research Conference, 2002. 60th DRC. Conference Digest. 2002:17–20.

Maloney, J.M., Uhland, S.A., Polito, B.F., Sheppard, N.F., Jr., Pelta, C.M., Santini, J.T., Jr. Electrothermally Activated Microchips for Implantable Drug Delivery and Biosensing. J. Control. Release. 2005; 109:244–255.

Meng, E., Chen, P.-J., Rodger, D.C., Tai, Y.-C., Humayun, M.S., Implantable Parylene Mems For Glaucoma Therapy. Proceedings of the 3rd Annual International IEEE EMBS Special Topic Conference on Microtechnologies in Medicine and Biology. 2005.

Musa, S., Welkenhuysen, M., Huys, R., Eberle, W., Kuyck, K.V., Bartic, C., Nuttin, B., Borghs, G. Planar 2D-Array Neural Probe for Deep Brain Stimulation and Recording (DBSR). IFMBE Proceedings. 2009; 22(21):2421–2425.

Najafi, K., Packaging of Implantable Microsystems. Proc. IEEE Sensors Conf. 2007.

Nguyen, N.-T., Huang, X., Chuan, T.K. MEMS-Micropumps: A Review. Transactions of the ASME. 2002; 124:384–392.

Osiander, R., Darrin, M.A.G., Champion, J.L. MEMS and Microstructures in Aerospace Applications. Taylor and Francis Group; 2006.

Prescott, J.H., Lipka, S., Baldwin, S., Sheppard, N.F., Jr., Maloney, J.M., Coppeta, J., Yomtov, B., Staples, M.A., Santini, J.T., Jr. Chronic, Programmed Polypeptide Delivery from an Implanted, Multireservoir Microchip Device. Nat. Biotechnol.. 2006; 24(4):437–438.

Rao, S.M.N., Mhatre, A., Popa, D.O., Chiao, J.C., Ativanichayaphong, T., Sin, J., Stephanou, H.E., MEMS-based Implantable Drug Delivery System. VII International Conference on Micro Electro Mechanical Systems. 2005.

Ratner, B.D., Hoffmann, A.S., Schoen, F.J., Lemons, J.E. Biomaterials Science: An Introduction to Materials in Medicine. San Diego: Academic Press; 1996.

Sassen, S., Voss, R., Schalk, J., Stenzel, E., Gleissner, T., Gruenberger, R., Neubauer, F., Ficker, W., Kupke, W., Bauer, K., Rose, M. Tuning Fork Silicon Angular Rate sensor with Enhanced Performance for Automotive Applications. Sensor. Actuator.. 2000; 83:80–84.

Schiff, L.I. Motion of a, Gyroscope according to Einstein’s Theory of Gravitation. Proc. Natl. Acad. Sci. USA. 1960; 46:871.

Shkel, A.M., Zeng, F.-G. An Electronic Prosthesis Mimicking the Dynamic Vestibular Function. Audiol. Neurotol.. 2006; 11:113–122.

Staples, M., Daniel, K., Cima, M.J., Langer, R. A Review of MEMS Drug Delivery in Medical Application. Pharm Res.. 2006; 23(5):847–863.

Tao, S.L., Popat, K.C., Norman, J.J., Desai, T.A. Surface Modification of SU-8 for Enhanced Biofunctionality and Nonfouling Properties. Langmuir. 2008; 24:2631–2636.

Thaysen, J., Yalcinkaya, A.D., Vettiger, P., Menon, A. Polymer-based Stress Sensor with Integrated Readout. J. Phys. D: Appl. Phys. 2002; 35:2698–2703.

Vasudev, A., Zhe, J. A Capillary Microgripper based on Electrowetting. Appl. Phys. Lett. 93, 2008. [103503].

Walker, J.A. The Future of MEMS in Telecommunications Networks. J. Micromech. Microeng.. 2000; 10:R1–R7.

Wang, J., Wise, K.D. A Hybrid Electrode Array With Built-In Position Sensors for an Implantable MEMS-Based Cochlear Prosthesis. J. Microelectromech. S.. 17(5), 2008.

Webster, T.J. Nanophase Ceramics:The Future Orthopedic And Dental Implant Material. In: Nanostructured Materials. New York: Academy Press; 2001:125–166.

Weinberg, M.S., Wall, C., Robertsson, J., O’Neil, E., Sienko, K., Fields, R. Tilt Determination in MEMS Inertial Vestibular Prosthesis. J. Biomech. Eng.. 2006; 128(6):943–956.

Whitesides, G.M. The Origins and the Future of Microfluidics. Nature. 2006; 442:368–373.

Wise, K.D., Wireless Implantable MicroSystems: Coming Breakthroughs in Health Care. International Electron Devices Meeting. 2006. [San Francisco].

Wu, J., Bernstein, G.H. A MEMS Inductive Deep Brain Stimulator. Sensors and Transducers (S&T e-Digest). 2006; 69(7):615–621.

Wu, J., Fedder, G.K., Carley, L.R. A Low-Noise Low-Offset Capacitive Sensing Amplifier for a 50 μg/√Hz Monolithic CMOS MEMS Accelerometer. J. Solid-State Circuits. 2004; 39(5):722–730.

Xie, H., Fedder, G.K. Fabrication, Characterization, and Analysis of a DRIE CMOS-MEMS Gyroscope. IEEE Sens. J.. 2003; 3(5):622–631.

Young, D.J., Development of Wireless Batteryless Implantable Blood Pressure EKG-Core Body Temperature Sensing Microsystem for Genetically Engineered Mice Real Time Monitoring. Proceedings of the 3rd International Conference on Nano/Molecular Medicine and Engineering. 2009. [Taiwan].

Zhou, J., Gu, Y., Fei, P., Mai, W., Gao, Y., Yang, R., Bao, G., Wang, Z.L. Flexible Piezotronic Strain Sensor. Nano Lett.. 2008; 8(9):3035–3040.

Zurcher, M.A., Young, D.J., Semaan, M., Megerian, C.A., Ko, W.H., MEMS Middle Ear Acoustic Sensor for a Fully Implantable Cochlear Prosthesis. Proc. 20th IEEE MEMS Conf. 2007:11–14.