Medical sensors in biomedical electronics, part 1: the eye and ear

Steve Taranovich - November 28, 2011

Almost 40 years after the popular sci-fi show The Six Million Dollar Man made its debut on TV, science fiction is becoming a reality as modern electronics technology merges with nanotechnology, advanced implants, solar- and light-powered devices, and major sensor advances in medicine and biology. Innovative marvels are transforming sensor-based electronics in human-body enhancements and replacements. These electronics include WBANs (wireless body-area networks) and enhancements and replacements for eyes and ears. Part one of this article describes innovative sensor technology and the miniaturized, implantable, and wireless-electronics-interface methods—from the sensor to the microcontroller. Part two discusses the lungs, heart, and brain.

The advanced developments in sensors and wireless-communications devices have enabled the design of miniature, cost-effective, and smart physiological sensor nodes. One innovation is the development of wearable health-monitoring systems, such as WBANs. The IEEE 802.15.4 standard for this technology stipulates a low-power, low-data-rate wireless approach in relation to medical-sensor body-area networks. STMicroelectronics this year contributed to this futuristic “cyborg” technology with its sensors and MEMS and the iNEMO (inertial-module-evaluation-board) node (Figure 1).

Medical sensors in biomedical electronics, part 1: the eye and ear figure 1

Medical sensors in biomedical electronics, part 1: the eye and ear figure 2

Among other vendors in this field, Analog Devices also offers some advanced activity-monitoring solutions and sensor-interface components, and Texas Instruments offers a development kit with Tmote Sky, a next-generation “mote,” or remote, platform for extremely low-power, high-data-rate sensor-network applications, which is designed with the dual goal of fault tolerance and development ease. TI’s Tmote Sky kit boasts a 10-kbyte on-chip RAM, the largest size of any mote; an IEEE 802.15.4 radio; and an integrated onboard antenna providing a range as far as 125m.

Helping the blind to see

Retinal-prosthesis development can help restore sight to people who have retinal-degenerative diseases, such as macular degeneration, which can cause blindness (Reference 1). Researchers doing preclinical implantation studies have verified that these prostheses ultimately assist an eye’s lost functions with an implant, containing a 15-channel stimulator chip, discrete power-supply components, and the power- and data-receiving coils that conform to the outer wall of an eye. In the study, researchers at the Boston Retinal Implant Project implanted an array in the subretinal area of a pig but affixed most of the prosthesis—a titanium, hermetically encased electronics assembly—to the outer surface of the sclera, or the white of the eye. A serpentine electrode array extends from the case to the superior temporal quadrant of the eye (Figure 2). The system has an external video-capture unit and a transmitter that wirelessly sends image data to the implanted portion of the device (Figure 3). A custom ASIC then translates the image into biphasic current pulses of Medical sensors in biomedical electronics, part 1: the eye and ear figure 3

Medical sensors in biomedical electronics, part 1: the eye and ear figure 3

programmable strength, duration, and frequency to the electrode array (Figure 4). Minco also offers advanced-design, flexible circuitry for implants that could help bring this project to reality for the approximately 1.7 million people who suffer from this eye condition.

Since the researchers performed this clinical study two years ago, many improvements in electronics technology have emerged that enhance miniaturization, decrease power consumption, and increase integration within this effort before it eventually becomes a product that the FDA (Food and Drug Administration) will approve for use on humans. Examples of these technological advancements include Texas Instruments’ Wireless Power Consortium Qi-compliant wireless receiver and transmitter technology. The company offers compliant communication for wireless-power transfer, ac/dc-power conversion, output-voltage conditioning, and dynamic-rectifier control for an improved load system. You can design complete contactless power transfer and charging with TI’s wireless-power products and development kits. Freescale and Analog Devices also offer low-power wireless products for this segment.

Medical sensors in biomedical electronics, part 1: the eye and ear figure 4

Another clinical study using photodiode circuits shows promise for the development of high-resolution retinal prostheses. In this study, researchers at Stanford University are investigating actively biased photoconductive and passive photovoltaic circuits (Reference 2). According to Daniel V Palanker, associate professor in the Department of Ophthalmology and in the Hansen Experimental Physics Laboratory at the university, a pocket PC processes a data stream from a videocamera, and a micro-LCD, similar to video goggles, displays the resulting images. An approximately 900-nm, nearly IR (infrared) light illuminates the LCD at 0.5-msec intervals, corresponding to approximately 30° of visual field. This pulsing projects the images through the eye optics into the retina. Photovoltaic pixels then receive the IR image in a subretinally implanted, 3-mm-diameter chip, corresponding to 10° of visual field. Each pixel converts the pulsed light into a proportionally pulsed biphasic electric current that introduces visual information into diseased retinal tissue.

The absence of an additional power supply in a photovoltaic system can greatly simplify prostheses design, fabrication, and the associated surgical procedures compared with photoconductive systems, which require an active bias voltage. The researchers plan future studies to determine the responses of the various retinal neurons to such stimulation.

Helping the deaf to hear

Medical sensors in biomedical electronics, part 1: the eye and ear figure 5

Another area of advancement in biomedical science covers cochlear implants. The primary goal of these implants is to use electrical stimulation safely to provide or restore functional hearing (Reference 3). The implants comprise a behind-the-ear processor in the external unit and a battery that uses a microphone to pick up sound, convert the sound to the digital realm, process and encode the digital signal into an RF signal, and then send it to the antenna in the headpiece (Figure 5). A magnet attracted to the internal receiver, which physicians surgically place just beneath the skin behind the ear, holds the headpiece in place. A hermetically sealed stimulator contains active electronic circuits that derive power from the RF signal, decode the signal, convert it into electric currents, and send them along wires threaded into the cochlea. The electrodes at the end of the wire stimulate the auditory nerve that connects to the central nervous system, which interprets electrical impulses as sound.

An external speech processor comprises a DSP, a power amplifier, and an RF transmitter. The DSP extracts features in the sound and converts them into a stream of data that the RF transmitter will transmit. The DSP also contains patient information in a memory map. An external-PC fitting program can set or modify the maps and other speech-processing parameters.

The internal unit has an RF receiver and a hermetically sealed stimulator. This internally implanted unit has no battery power, so the stimulator must derive its power from the RF signal. The charged stimulator then decodes the RF bit stream and converts it into electric currents for delivery to appropriate electrodes at the auditory nerve. A feedback system monitors critical electrical and neural activities in the implants and transmits these activities back to the external unit (Figure 6).

Medical sensors in biomedical electronics, part 1: the eye and ear figure 6

Advanced Bionics has developed an implantable electronics platform that benefits patients by offering more channels and the ability to generate virtual channels through current steering. According to Lee Hartley, vice president of R&D at the company, one of the biggest challenges in developing sophisticated sound-processing sensors is improving the ability to hear in noisy listening environments. “Cochlear-implant recipients have a reduced ability to discriminate loudness levels and distinct frequency channels,” he says. “This [reduced ability] heightens the challenge of improving speech understanding and music appreciation; we need to intelligently separate information from noise.”

The next major areas for significantly improving cochlear-implant systems and performance, says Hartley, include ubiquitous wireless connectivity to commercial devices, increasingly intelligent scene-analysis algorithms running at low power, and technologies that enable patients to receive cochlear-implant services from clinicians regardless of the patients’ or clinicians’ location. “Technology trends in the industry are moving toward system architectures and service models that will minimize the visibility of the entire cochlear-implant system,” he explains. Hartley expects advances in IC technology to afford the delivery of wireless features and system-power reductions: “I see system design continuing to be modular in that recipients will customize their experience based on their changing needs.”

Signal processing has greatly improved the performance of cochlear implants. Sound can be modeled either as a periodical source for voice sounds or as a noise source for unvoiced sounds. The resonance properties in the vocal tract filter the sounds’ frequency spectrum. Alternatively, the source can be modeled as a carrier while the vocal tract acts as a modulator, reflecting the opening and closing of the mouth or the nose. The source typically varies rapidly, whereas the filters react more slowly (Reference 3).

The internal unit in all modern cochlear implants connects to the external unit by a transcutaneous RF link for the safety and convenience of the user. The RF link uses a pair of inductively coupled coils to transmit not only data but also power. The RF-transmission unit has some challenging tasks, such as efficiently amplifying signals and power and maintaining immunity to EMI. Its secondary functions are to provide reliable communication protocols, including a signal-modulation method; bit coding; frame coding; synchronization; and back-telemetry detection.

The RF design of cochlear implants presents many conflicting challenges that require careful compromises. For example, to extend battery life, the power transmitter must be a high-power, efficient design. Thus, most modern implants use a highly efficient Class E amplifier. Class E amplifiers are nonlinear, however, and their distorted waveform limits the data-transmission rate. Another challenge is the need for power-efficient transmitting and receiving coils. Operating the RF system at its resonant frequency, or at a narrow bandwidth, maximizes power, but the RF system must have unlimited bandwidth for data transmission. And, although these devices call for high transmission frequency, this requirement dictates a large coil. In a practical, usable design, however, the size of the transmitting and the receiving coils must be small and cosmetically acceptable.

The receiver and stimulator in the internal unit act as the engine of the cochlear implant (Figure 7). The ASIC (shown in dashed box) performs the critical function of ensuring safe and reliable electrical stimulation. It has a forward pathway with a data decoder that recovers the digital information from the RF signal, an error and safety check that ensures proper decoding, and a data distributor that sends the decoded electrical-stimulation parameters to the programmable current source by switching the multiplexers on and off. The backward pathway includes a back-telemetry voltage sampler that reads the voltage for a time on the recording electrode. The PGA (programmable-gain amplifier) then amplifies voltage, the ADC converts it to the digital domain and stores it in memory, and the back-telemetry technology sends it to the external unit. The ASIC also has many control units, which range from the RF signal generated from the clock to the command decoder. The ASIC cannot easily integrate some functions, such as the voltage regulator, the power generator, the coil and RF-tuning tank, and the back-telemetry data modulator, but advances are occurring in these areas.

Medical sensors in biomedical electronics, part 1: the eye and ear figure 7

The current-source circuit, comprising a DAC and current mirrors, generates the stimulating current according to the amplitude information from the data decoder. This current source must be accurate and involves challenges. For example, due to process variations, the relationship between the source and the drain of the MOSFET is not constant, yet the voltage difference between the gate and the source controls the amount of current in the drain. For this reason, the circuit requires a trimmer network to fine-tune the reference current. New designs combine multiple DACs to obtain the desired accurate current, thereby eliminating the need for a trimmer. An ideal current source also has infinite impedance, so some designers use cascoded current mirrors at the expense of reduced voltage compliance and increased power dissipation.

You must carefully consider and implement these compromises. Some cochlear-implant products have multiple current sources, and older devices required a switching network to connect one current source to multiple electrodes. Recent designs use multiple current sources sequentially or simultaneously, however. In these designs, both the P- and the N-channel current sources generate positive and negative phases of stimulation. The challenge is to match the P- and the N-channel current sources to ensure balancing of the positive and the negative charges. Adaptive compliance voltages can reduce power consumption and maintain high impedance.

Engineers prefer ASK (amplitude-shift-keying) modulation over FSK (frequency-shift-keying) modulation because of ASK’s simple implementation scheme and low power consumption with the high-frequency RF signal. Thanks to persistent and collaborative work by teams of engineers, scientists, physicians, and entrepreneurs, safe and charge-balanced stimulation has restored hearing to more than 120,000 people worldwide. These prostheses serve as models to guide development of other neural prostheses to improve the quality of life for millions of people.

You can reach Contributing Technical Editor Steve Taranovich at staranovich@yahoo.com.

This article is the first in a series on medical electronics. Read part 2: “Medical sensors in biomedical electronics, part 2: the brain, heart and lung."

References

Kelly, Shawn K; Douglas B Shire; Jinghua Chen; Patrick Doyle; Marcus D Gingerich; Stuart F Cogan; William A Drohan; Sonny Behan; Luke Theogarajan; John L Wyatt; and Joseph F Rizzo III, “A hermetic wireless subretinal neurostimulator for vision prostheses,” IEEE Transactions on Biomedical Engineering, Volume 58, No. 11, November 2011.
Loudin, James D; Stuart F Cogan; Keith Mathieson; Alexander Sher; and Daniel V Palanker, “Photodiode Circuits for Retinal Prostheses,” IEEE Transactions on Biomedical Circuits and Systems, Volume 5, No. 5, October 2011, pg 468.
Zeng, Fan-Gang; Stephen Rebscher; William Harrison; Xiaoan Sun; and Haihong Feng, “Cochlear Implants: System Design, Integration, and Evaluation,” IEEE Reviews in Biomedical Engineering, Volume 1, 2008, pg 115.

For More Information
Advanced Bionics	Analog Devices	Boston Retinal Implant Project	Freescale
Minco	Stanford University	STMicroelectronics	Texas Instruments
VA Boston Healthcare System	Wireless Power Consortium