U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Michael AC, Borland LM, editors. Electrochemical Methods for Neuroscience. Boca Raton (FL): CRC Press/Taylor & Francis; 2007.

Cover of Electrochemical Methods for Neuroscience

Electrochemical Methods for Neuroscience.

Show details

Chapter 20Telemetry for Biosensor Systems

and .

Introduction

This chapter will provide an overview of radio frequency (RF) telemetry systems by examining the design requirements of two contrasting systems: a wireless system designed to support Fast Scan Cyclic Voltammetry (FSCV) and a wireless system designed to support selective biosensors. Traditionally, RF design and analog circuit design were very complicated endeavors. Component interactions and non-ideal behaviors made such designs as much art as science. Today however, there are a wide variety of analog integrated circuits with near ideal behavior, and self-contained RF modules with built-in standardized protocols are available. In short, it is now very possible for a neuroscientist, with little or no knowledge of telemetry, to design and construct robust wireless acquisition systems.

In addition, there are a wide variety of commercial implants available for identification and physiological monitoring in animals. Currently available wireless implants can measure activity, core temperature, blood pressure, heart rate, blood flow, pH, biopotentials Electrocardiography (ECG), Electroencephalography (EEG), Electromyography (EMG) and respiratory rate [1]. Major suppliers include Data Sciences International (www.datasci.com), Biomedic Data Systems (www.bmds.com), and Respironics (www.minimitter.com). Most of these systems use some form of inductive power and telemetry, very low power transmitters or both. That is, these devices normally use proprietary low power analog transmitters that can only communicate over short distances, and in some cases the power required to operate the device is derived from RF energy transmitted from the receiver (Reader) to the device. A wide and impressive variety of Application Specific Integrated Circuits (ASIC) has also been developed. Mohensi et al. provide an excellent overview of current systems [2].

The goal of this chapter is simply to provide an overview based on familiar examples in the hope that this will help the reader to make informed decisions, whether building or buying telemetry based systems. To that end, we will consider the design of a digital wireless FSCV system for the measurement of electroactive species such as dopamine, and a biosensor based digital wireless system.

Fast Scan Cyclic Voltammetry

FSCV has become the standard tool for neuroscientists to study the kinetics of neurotransmitter release and uptake in the rat brain [3]. In vivo cyclic voltammetry is a powerful technique that allows the direct measurement and identification of compounds in the brain. Typically, FSCV for neurotransmitter applications uses scan rates around 300 V/s repeated 10 times per second. FSCV, in combination with very small carbon fiber electrodes is used for a number of reasons. Microelectrodes (6–30 μm diameter) provide superior spatial resolution from a group of synapses, however use of microelectrodes requires very fast scan rates due to the diffusion characteristics and kinetics of the electrochemical reaction [4,5]. FSCV provides a unique “fingerprint” for specific electroactive species, which serves to increase selectivity from potential interferences. The small electrode size creates a very small diffusion layer, thus leading to enhanced spatial and temporal resolution. In addition, the amount of toxic electrogenerated species will be quite small, preserving the integrity of the measured environment [6]. One limitation of FSCV is high background. As a result of the double layer capacitance, a large background current is always observed. Background currents can be easily identified in vitro. Wightman’s group has described methods by which average background current can be acquired in vivo. In general, this involves acquiring multiple scans while biological activity is minimal and averaging them [7].

FSCV is widely used for the in vivo measurement of dopamine [8–11]. While other techniques, such as microdialysis, can be used for this application, FSCV has a number of advantages, as outlined above. In addition to superior spatial resolution, FSCV also has better temporal resolution. Speed is essential because neurotransmitter (such as dopamine) activity tends to occur in phasic bursts that only last for 1–2 s [11]. Microdialysis probes have also been shown to produce damage that alters measurements near the probe [12].

FSCV System Requirements

Typical FSCV system requirements are:

  • Three electrodes (1 working, 1 reference, and 1 counter)
  • User-controlled sweep repeat rate
  • User-controlled scan rate (0–1000 V/s)
  • User-controlled voltage range of − 1.0 V to 2.0 V
  • User-controlled Low Pass Filters
  • Total samples per sweep of 1000
  • At least 12-bit analog output resolution
  • At least 14-bit analog input resolution (≥ 100 kHz sampling rate).

From this list the telemetry data rate requirements can be estimated. Assuming a repetition rate of 10 scans per second and a total sweep time of less than 30 ms, 1000 14-bit data points will be acquired every 100 ms, and there will be a 70 ms window in which to transmit these data between scans. Given processing overhead it is prudent to assume that only 50 ms will be available for transmission. Further assuming a 2-byte per sample transmission, the minimum data rate required is 16 (bits) * 1000 (Samples/Sweep)/0.05 (seconds/Transmission) = 320 kbits/s or 40 kBytes/s. This estimate does not include compression or packetization/balancing overhead. FSCV also requires a significant amount of set-up (scan rate, scan type, upper and lower bias, repeat rate, gain, stimulus). This indicates the need for a 2-way telemetry scheme. That is, a scheme in which data can be both transmitted and received.

In-Vivo Electrochemical Sensors

The pioneering work of Adams [13,14] and Wightman [15] on in vivo electrochemistry has demonstrated the utility of microsensors in the study of important issues in neurophysiology and neuropharmacology. These devices are generally confined to species that have well-behaved direct electrochemistry such as dopamine.

By definition, an electrochemical biosensor is “a self-contained integrated device, which is capable of providing specific quantitative or semi-quantitative analytical information using a biological recognition element (biochemical receptor) which is retained in direct spatial contact with an electrochemical transduction element.” [16] In general, enzymes are used to selectively convert the substrate of interest into an equivalent amount of a readily detected electroactive species. Microbiosensors offer high selectivity, spatial, and temporal resolution, and considerable progress has been made in their design and application [17,18]. A recent review provides an excellent overview of the current state-of-the-art [19]. Biosensors have been developed for the detection of wide range of analytes including glutamate, choline and acetylcholine, lactate, pyruvate, and of course, glucose. Figure 20.1 shows an illustration of a glutamate biosensor available from Pinnacle Technology, Inc. Lawrence, KS.

FIGURE 20.1. Glutamate biosensor overview (Pinnacle technology).

FIGURE 20.1

Glutamate biosensor overview (Pinnacle technology).

The electronics support system required for biosensors is quite simple. For amperometric biosensors, a standard potentiostat circuit is used to maintain a fixed voltage while measuring a very small signal current. Biosensors are selective by design, and their temporal response is typically longer than 50 ms. This makes the design of a telemetry system much simpler than it is for FSCV.

Biosensor System Requirements

These specifications will change with the type of biosensor targeted, but a typical set of biosensor system requirements is:

  • 2 or 3 electrodes
  • Static voltage (cannot be removed)
  • User controlled bias range of (0–1.3 V)
  • Gain of 20 million V/A
  • Sample rate of 1 sample/second
  • At least 12 bits of analog output resolution
  • At least 16 bits of analog input resolution.

The telemetry and power requirements in this case are much lower than those required for FSCV. Even allowing a factor of two safety margin for various overhead parameters and a couple of channels, the required data rate is 2 (safety) * 2 (channels) * 1 (sample/seconds) * 16 (bits/sample) = 64 bits/s or 8 Bytes/s. The number of changeable parameters in this type of system is also low. That is, it may be enough to allow bias to be changed via a wired interface. In this case, a one-way (transmit only) telemetry system may be employed.

General Design Considerations

Wireless systems offer a number of advantages, the primary two being freedom of movement for the animal under examination and elimination of all cable artifacts. There are, of course, associated drawbacks. In particular, these systems require that all functional blocks including power, telemetry, data acquisition, and conditioning, are mounted on or in the animal.

As a rule, any system developed should relieve stress for both the animal under test and the researcher carrying out the experiment. To this end, it is very important for system designers to work with end users to optimize the mechanical aspects of the system (reduced surgery time, simple sensor placement, simple battery replacement/recharging/recalibration), and software (simple, intuitive graphical user interfaces, reliable telemetry, reliable, and robust data storage).

Another obvious, but very important consideration for external systems designed for use with rats or mice, is that both are very small. Animal testing at Northwestern University has shown that mice will tolerate a head-mounted unit of up to 1 cm [3] without modifying their running wheel behavior, as long as the weight does not exceed 0.5 g (E. Naylor, personal communication). Others report upper weight limits for head mounted devices as 25 g for rats and 5 g for mice [20]. This presents a significant problem, and for this reason, most wireless systems produced for mice are internal implants. In the discussion that follows, we will not make any distinction between external systems (head mount, back mount) and implants since the operating principles are very similar. For rats there is an additional concern. Rats are quite aggressive in the removal of any external systems, so any external monitoring system developed for rats needs to be well protected.

For a wireless system, the designer has the choice of transmitting the data acquired in an analog form, or of digitizing on board before transmitting. Frequency modulated analog transmission can be relatively noise free, but the resolution of such systems is constrained by the resolution of the frequency modulation electronics on the transmitter, and the demodulator on the receiver. As was previously mentioned, there are a number of ASICs that have been developed for this purpose, but in general, it is preferable to digitize at the point of measurement if possible. In this case, the resolution and accuracy of the system are under direct control of the sensor electronics, and once the signal is digitized, no resolution is lost in the telemetry regardless of type.

Electronics

Batteries and other power sources will be discussed in a subsequent section, but it is important to note that when calculating battery life, the average current consumed is the key parameter—not the peak current. It will often be desirable to use a high data rate link in order to minimize the amount of time that the transmitter is on. Typically, the RF transmitter will consume more power than any other part of the system, so minimizing the amount of time that the telemetry system is on minimizes power consumption. FCC licensing is also simplified if the duty cycle (on time/total time) is less than 1%. In some situations, a fast, high-power telemetry system, on average, may require less power than a slow, low-power telemetry system.

A digital wireless data acquisition system will normally contain a small, low-power controller of some type, as well as analog to digital converters (A/D), digital to analog converters (D/A), sensor conditioning, power and telemetry. Battery voltage is typically limited by the voltage requirements of the controller (usually 2.7 V), and headroom for the amplifiers.

Assuming that the system being developed will digitize the data onboard, then the A/D type and range must also be carefully evaluated. For many low frequency and high-resolution applications sigma-delta A/Ds are very attractive. A/Ds of this type basically trade bit-resolution for sample speed. That is, they can be extremely accurate if the sampling rate is low. They are capable of measuring low-level signals directly and they normally implement an internal filter that can remove, or reduce the size of required anti-aliasing filters.

In many cases, this type of system can be designed to run directly off a voltage reference. References come in a variety of forms, but they are basically just Zener diodes held in breakdown with some form of buffered output. Voltage references provide a very stable, known voltage that ensures that the A/D bit measurement can be accurately translated to voltage. A linear voltage regulator may also be required. If needed, a regulator should be employed that works when the input voltage is very close to the desired voltage (low drop out), and for highest efficiency, the systems should be operated as close to the drop-out point as possible. Switching regulators should be avoided for most designs of this type. Switching regulators effectively operate by changing a DC input to an AC intermediary and then back to a DC output at the same or different voltage. These devices feature low cost, small size and low parts count, but they typically are not suitable for low noise and low power applications.

Power Sources

New power sources are currently attracting considerable attention. These include advanced battery designs, microfuel cells, biofuel cells, and energy harvesters that convert light or vibration to energy. In the future, one or more of these solutions may become valuable for wireless design, but at present, they either target incompatible primary markets such as hybrid vehicles or cell phones, they are not commercially available at a reasonable price or they do not produce enough power to be useful in a data streaming system. For most research applications, it is preferable to use a power solution that is readily available and inexpensive. For this reason, we will focus on primary (not rechargeable) and secondary (rechargeable) batteries. Note: milliamp-hour (mAh) energy specifications are often optimistic, and normally involve discharge to a voltage level that may be significantly less than the minimum operating voltage of the system.

There are a large number of battery chemistries to choose from, however, if one of the system design parameters is small size, then the options are very limited. Battery availability is generally driven by the laptop, cell phone, hearing aid, and game markets. For many animal and human telemetry designs the main concerns are size, weight, capacity, discharge profile, and pulse current capability. In general, size and weight are a function of energy density. That is, how much energy can a given battery chemistry store in a given weight and volume. However, be cautious of data sheet claims. Practical energy density, including enclosure materials, electrolytes, etc., are often significantly lower than the theoretical energy density that is calculated based solely on the properties of the active materials. In addition, the chemical reaction in a primary cell will slow down at low temperatures degrading their ability to supply energy. At high temperatures, self-discharge can become a problem. For animal research size and weight are normally the most important parameters.

Primary batteries can be subdivided into five main classes: zinc/carbon/ammonium chloride (dry cell), zinc-air, silver oxide, alkaline, and lithium. Of these, the silver oxide and lithium chemistries are most useful for small wireless systems. Silver oxide button cells offer reasonable peak currents and a flat discharge curve. Lithium generally supports lower discharge currents, however lithium offers light-weight, high energy density, high open circuit voltage, low self-discharge, a flat discharge curve and a variety of small form factors. Three primary lithium chemistries are of specific interest, lithium carbon monofluoride (LiCFx), lithium manganese dioxide (LiMnO2), and lithium thionyl chloride (LiSOCl2). Zn-Air energy density and storage life are also excellent and this chemistry may be a good choice for some systems. Table 20.1 gives an overview of the general characteristics of several important primary battery types. Pulse capability describes the ability of a battery to deliver pulse currents. This is important in RF systems since relatively high pulse currents may be required when transmitting and receiving data. Flat discharge describes the discharge characteristics of a battery under load. A flat discharge profile is desirable in most cases since the battery will maintain a steady voltage throughout the bulk of its discharge, and then the voltage will fall quickly. This makes it easier to design systems that use more of the available energy in the battery.

TABLE 20.1

TABLE 20.1

Primary Battery Comparison by Type

Rechargeable batteries generally offer low energy densities, and they require support electronics to implement the correct charging protocols. In the case of rechargeable lithium cells, special protection circuitry may also be required to ensure user safety. The cell voltage of rechargeable batteries also tends to be lower. However, for many applications, the convenience of recharging by direct, inductive or optical means may make a rechargeable chemistry preferable. For very small footprint devices, useful chemistries include: nickel cadmium (NiCd), nickel metal hydride (NiMH), and lithium-ion. Memory effects in Ni chemistries due to crystal formation are not normally a problem with modern battery designs. A wide range of lithium-ion and lithium-ion polymer (solid electrolyte) chemistries are available. The construction of rechargeable lithium cells from different manufacturers can be quite different. This makes it difficult to summarize the behavior. Table 20.2 shows the general characteristics of two of the most common materials. It is also worth noting that it is possible to form thin film lithium-ion batteries that can be molded into the mechanical structure of a device.

TABLE 20.2

TABLE 20.2

Secondary Battery Comparison by Type

Specialty suppliers such as Wilson-Greatbatch (www.greatbatch.com) and Quallion (www.quallion.com) offer a wide variety of solutions targeting medical implants, and some of these can be useful for wireless designs. For higher power and high pulse applications, Tadiran (www.tadiran.com) offers a wide range of lithium solutions. Another option that should be considered is supercapacitors. Supercapacitors offer the possibility of very fast charging, high discharge currents, light-weight, and low cost. As more small format options become available, supercapacitors may come to play a greater role in wireless designs.

Rf Telemetry

This section will focus on devices that operate in the Industrial Scientific and Medical (ISM) band of the RF spectrum. The ISM band includes 315–433 MHz, 902–928 MHz, 2.402–2.480 GHz, 5.725–5.875 GHz, and 24.0–24.25 GHz. In general, the maximum data rate is lower at lower frequencies. Telemetry in these bands is unlicensed in the U.S., but specific FCC regulations apply to the various bands. European regulations differ from those in the U.S., so careful attention must be paid to the various regulatory requirements if a device is intended to operate internationally. One drawback of the unlicensed ISM band is that it suffers from the “Tragedy of the Commons.” [21] That is, use within these bands is minimally restricted and in some cases abused, so freedom from interference can never be guaranteed. A current example is the performance degradation and sometimes complete failure of wireless datalinks from laptop computers, wireless cell phones, etc., all vying for bandwidth in the very crowded 2.4 GHz band. Depending on the application, the Medical Implant Communications Service MICS (402–405 MHz), and the Wireless Medical Telemetry Services (608–614 MHz) may also be of interest.

True discrete RF circuit design requires considerable expertise. In most cases, it is preferable to buy and implement a module than to implement any solution from scratch. If you decide to pursue a discrete solution, follow the manufacturer’s circuit and layout guidelines exactly . The following sections will discuss some of the currently available RF module and protocol options.

Protocol Independent Transmitters and Receivers

The simplest telemetry systems require nothing more than an inductor/capacitor (LC) circuit and an amplifier. These systems are very easy to construct, but they are relatively inefficient and subject to drift based on proximity to ground planes, temperature, etc. Transmitters that are much more stable can be constructed from crystals, phase locked loops (PLLs) or surface acoustic wave (SAW) elements. A variety of vendors (RF Monolithics www.rfm.com, Linx Technologies www.linxtechnologies.com, etc.) offer prefabricated and precertified transmit and receive modules that require little more than a data input/output, and antenna connection to set up a simple wireless link. Transceiver modules are also readily available. The data rates of these modules tend to be low (< 200 kbits/s), but for simple systems, they can be very useful. One drawback to solutions of this type is that they are normally fixed to one frequency. This can create problems when multiple devices are in range of one another. For low duty cycle transmitters, simple time domain multiplexing (TDM) and error correction schemes can be used to overcome these limitations.

Zigbee (Ieee 802.15.4)

The ZigBee (IEEE 802.15.4) protocol offers reasonable data rates (250 kbit/s at 2.4 GHz), low power, small size, and mesh capable operations. ZigBee devices are specifically targeted to very low power, battery operated applications where the device itself is expected to be off 99.9% of the time. ZigBee is also self-assembling. Up to 65,000 devices can auto-assemble in a mesh with a single coordinating node. ZigBee is also “self-healing” in that it can detour data around malfunctioning nodes. ZigBee can operate in three bands for international compliance (868 MHz—1 channel, 915 MHz—10 Channels, 1.4 GHz—16 channels). Major module vendors include, Crossbow (www.xbow.com), Ember (www.ember.com) and Maxstream (www.maxstream.com). ZigBee modules typically require approximately 25–50 mA at 3 V when operating, but the current consumption in shutdown is very low (<25 uA). The power requirements are dependent on the desired range.

Bluetooth

In its base form, Bluetooth provides a two-way data link at up to 721 kbits/s and 3 voice channels for distances up to 100 m. The version 2.0+ EDR specification extends the data rate to 3 Mbit/s. Bluetooth is a robust specification operating in the 2.4 GHz ISM band. Bluetooth specifies three separate power ranges depending on the range required: Class 1 (100 m), Class 2 (10 m) and Class 3 (10 cm). Bluetooth normally requires that a user manually pair devices. The primary markets for Bluetooth devices are in local device-to-device communications at moderate data rates—primarily wireless headsets for phones and audio equipment. A class 2 Bluetooth device typically requires 25–50 mA when operating and 1–2 mA when in idle mode. The power requirements are dependent on desired range. Bluetooth modules are available from a wide range of suppliers.

Wifi (Ieee 802.11)

Of the various incarnations of 802.11, 802.11b at 11 Mbit/s is most commonly used in embedded systems, but 802.11 g at 54 Mbit/s can be easily implemented. WiFi targets wireless computer networking and as such it is typically overkill for sensor systems. However, there are applications that require the robust infrastructure and high data rates offered by this protocol. Unfortunately, in this case, speed requires power. WiFi modules will normally consume about 1 watt when operating. WiFi modules for embedded systems are becoming commonplace (www.lantronix.com, www.digi.com).

Ultra Wideband

Ultra Wideband (UWB) is a very interesting new protocol that targets very high speed 100– 480 Mbit/s, short-range (10–20 m) video and high-speed data applications. The basic technology uses 500 MHz channels in a 7 GHz range (3.1–10.6 GHz). By using very wide bands, the power at each frequency can be very low, so this technology is not likely to interfere with other devices. The wide bands also preclude interference from other technologies that may be using the same bands. The power required by these devices is also very low, and the module size is small, so in the future UWB should be very useful for a wide range of data streaming applications. At present, consumer devices such as wireless USB 2.0 and Bluetooth links over UWB are just being introduced.

Radio Frequency Identification

Radio Frequency Identification (RFID) is being developed to replace UPC bar codes. RFID devices may be passive or active. Passive devices derive the power from the electric or magnetic field generated by a reader, and respond to the reader by detuning a tuned circuit. From the reader’s point of view, less energy is coupled to the RFID device when it is detuned and this can be interpreted as binary data. The typical range of passive RFID tags is about 15 ft. (dependent on frequency). Data rates are very dependent on the carrier frequency and type of data modulation used. Amplitude shift key (ASK), frequency shift key (FSK), and phase shift key (PSK), are common approaches. For example, in an FSK scheme, the tuned circuit may be tuned and detuned for 4 and 4 carrier cycles to represent a binary 0, and 5 and 5 carrier cycles to represent a binary 1 with resulting data rates in the range of 30 kbits/s. Active RFID devices typically contain a battery that drives a transmitter in response to an RF query. These devices can operate over much longer distances. A common example is the devices used for automatic toll collection on many highways. RFID devices operate in a wide range of bands (125–135 kHz, 13.56 MHz, 868–928 MHz, 2.45 GHz, 5.8 GHz). RFID devices can be, and have been, applied to sensor applications, however, there are two main problems. The first is that the driving markets for RFID do not require data acquisition of any kind. They just return a unique identification in response to a query. For this reason, the bulk of the devices that are designed specifically for RFID have no acquisition capability. The second problem is that many of the techniques relevant to neuroscience require very accurate measurement of low-level signals. The RF interrogation and power coupling from an RFID reader generates a substantial amount of electrical noise which can severely degrade the signal to noise ratio if the system is not very carefully designed. The amount of power that can be coupled in is also very limited. Another problem unique to biosensors is that the applied potential on the sensor must be maintained. This can be achieved with RFID only if the on-board capacitance is kept sufficiently high.

Interference

In a normal operating environment, many different devices may be sharing the same RF band, and non-intentional radiators such as motors and microwave ovens can contribute a significant amount of RF interference in a band. This is particularly true in the ISM bands. Transmission in these bands is governed by the Federal Communication Commission’s FR47 Part 15 rules. These rules place significant restrictions on power and modulation format. The simplest RF systems use only a single carrier frequency. Very low power transmitters can be located in close proximity as long as the transmitter range is small in comparison to the unit separation. In many cases, higher power single frequency systems can also be closely spaced if the duty cycle of the transmitter and receiver is very low and each unit contains a unique digital identification. More advanced systems such as ZigBee, Bluetooth and WiFi use spread spectrum techniques to overcome these limitations. As an example, Bluetooth implements frequency hopping spread spectrum modulation (FHSS) changing between 79 different randomly chosen frequencies in the 2.402–2.48 GHz range 1600 times each second. ZigBee and WiFi use a related technique called direct sequence spread spectrum in which the data are transmitted over a broad range of frequencies. In this way, a single frequency interference source cannot disrupt communications although broad frequency non-intentional noise sources can, and as an added benefit, the signals are generally “noiselike” and therefore, they are difficult to intercept. Hybrid modulation formats such as distributed frequency spread spectrum (DFSS) attempt to combine the benefits of the FHSS and DSSS.

Mesh Networks

Although not necessary for most neuroscience applications, it is worth noting that ZigBee, Bluetooth, and WiFi can all be operated in a mesh network architecture. With this structure, each unit communicates with its nearest neighbor until the data are finally transferred to/from a central controller. Systems of this type are robust since the loss of one node in the network will not affect the overall signal flow; they are simple to implement, they can use very little power since the transmitter only needs to be powerful enough to communicate with its closest neighbors, and they can work well in difficult RF environments since multi-path effects are greatly reduced.

Antenna Design

Regardless of telemetry type, an antenna will be needed. The antenna must be small enough to meet the requirements of a particular application, but it must also be matched to the RF system. That is, in the case of a transmitter, power flowing from the transmitter should couple to the antenna without reflection. In general, matching an antenna to a particular RF system will require consultation with someone who has the necessary tools and expertise to evaluate the design. In typical small medical systems everything in the system contributes to the RF circuit. This can include the particular size and type of battery being used, thickness of substrate, available ground plane, proximity of active components, and proximity of other ground planes (i.e., the animal itself), etc. That said, it is possible to get a good solution by trial and error. Most vendors of RF systems will provide on-line antenna design guides that are an excellent starting point. Beyond that, most small antenna designs will look capacitive, so in the case of a transmitter, a reasonable approach is to place a range of inductors in series with the transmitter and monitor the signal strength at a known receiver. When the inductor that provides the highest signal strength or transmit distance is found, the design is probably not perfect, but it may be adequate. Also, note that all small antennas will be directional to some extent. If all orientations are possible in an experimental situation, it is important to specify the range of a newly designed RF system in its worst possible orientation.

Any antenna designed for animal use should also be comfortable for the animal. The frequencies and levels of RF energy under discussion are both non-ionizing and non-thermal (no local heating). There is still some debate about possible health issues related to cell phones, but the power output of cell phones is much higher than any of the RF systems under discussion. In short, there is no evidence that exposure to digital RF energy at these levels will have any adverse effects on the animals under test.

IR Telemetry

Infrared (IR) telemetry should also be considered for any external design. IR telemetry systems can be very small, require no antenna, and although they require a considerable amount of power to operate, high data rates are possible, and as mentioned previously, this means that a capacitor can be used to create a reservoir of charge to supply power for a short period of time. The average power draw can be very small. Standards set by the Infrared Data Association specify 16 Mbit/s transmission rates currently with 100 Mbit/s standards planned. The primary drawback for IR telemetry is that it is line of sight. It is possible to use multiple receivers to eliminate this problem to some extent, but in general, IR is not appropriate for small animals that may sleep curled up, or under bedding.

Design Examples

FSCV

With size and power consumption in mind, it is likely that the computational power available on a wireless FSCV system will be quite small. However, computationally simple Huffman coding techniques in conjunction with derivative computation have been shown to deliver < 50% compression on EEG signals [22]. When all of these requirements are taken into account, the end result is that using available technology, a wireless FSCV system will require a 2-way data link capable of at least 200,000 bits per second (bit/s) and the overall system will consume approximately 0.1 watts of power.

Of the RF solutions discussed, only Bluetooth and UWB are appropriate. At this time, there are many more Bluetooth embedded solutions available, so this would be the most reasonable choice. The power consumption of this system will be considerable and the duty cycle will be high, so a relatively large power source will be required. Of those discussed, two LiSOCl2 half AA batteries at 3.6 V may be ideal. With a capacity of approximately 1.5 Ah, this could be used to power the system for a few days of continuous use. Secondary lithium cells would also be appropriate, but for most research use, the cost of primary cells will be insignificant, and the added battery lifetime will be valuable.

Selective Amperometric Biosensors

As described the data rate and power requirements of amperometric biosensor telemetry systems is quite low. This offers a range of RF and power solutions that will work well. ZigBee is a very reasonable choice since the duty cycle of this type of system will be very low. That is, data is transmitted in very short bursts with a long period of inactivity between transmissions. ZigBee offers two-way secure communications at very low power levels, and the ability to place numerous devices in close proximity with no interference, however, there are also some disadvantages to this choice: ZigBee module sizes are relatively large, and ZigBee requires a significant amount of power while transmitting and receiving. Typical lithium coin cells cannot deliver a pulse of more than approximately 1 mA without experiencing a severe voltage drop. Other small button cells such as silver oxide and zinc air can deliver the required pulse currents without a large capacitive assist, but the cell voltages are low, so multiple batteries may be required to implement a solution.

As an alternative, a simple transmission only solution using any of the available 315, 433 or 915 MHz (868 MHz for Europe) modules may be implemented. The advantage to this type of solution is that lithium coin cells with a capacitive assist can be used to implement a very small, light solution. There are several disadvantages. The first is that this type of solution uses a fixed frequency and therefore is more susceptible to interference from other systems and noise sources. It is also not possible to communicate commands to the device. That is, it is not possible to request a gain or bias change without some form of secondary wired interface. This type of solution will also have lower data rates than a ZigBee based solution. The result is that although the transmitter draws less power when on, the transmitter may be on for a much longer time. A final decision requires a full energy analysis for the specific application and a full evaluation of user requirements.

Conclusions

The purpose of this chapter was to present an overview of the design of wireless systems that is appropriate for in vivo measurements. The general topic is very broad, and many details have been omitted for the sake of brevity. The gist of the argument is that a great deal of industrial research has targeted the design of small, plug-and-play telemetry modules, and with these and some general knowledge of electronics, any neuroscientist can evaluate and design wireless data acquisition systems.

References

1.
Kramer K, Kinter L. Evaluation and application of radio telemetry in small laboratory animals. Physiol Genomics. 2003;13:197–205. [PubMed: 12746464]
2.
Mohensi P, Naja K, Eliades S, Wang X. Wireless multichannel biopotential recording using an integrated FM telemetry circuit. IEEE Trans Neural Syst Rehabil Eng. 2005;13:263–271. [PubMed: 16200750]
3.
Robinson D, Venton BJ, Heien MLAV, Wightman RM. Detecting subsecond dopamine release with fast-scan cyclic voltammetry in vivo. Clin Chem. 2003;10:1763–1773. [PubMed: 14500617]
4.
Bard AJ, Faulkner LR. Electrochemical Methods; Fundamentals and Applications. Wiley; New York: 2001.
5.
Brett CMA, Brett AMO. Electrochemistry Principles, Methods, and Applications. Oxford University Press; Oxford; 2000.
6.
Zimmerman JB, Wightman RM. Simultaneous electrochemical measurements of oxygen and dopamine. Anal Chem. 1991;63:24–28. [PubMed: 1810167]
7.
Michael D, Joseph J, Kilpatrick M, Travis E, Wightman RM. Improving data acquisition for fast-scan cyclic voltammetry. Anal Chem. 1999;71:3941–3947. [PubMed: 10500480]
8.
Jones SR, Garris PA, Kilts CD, Wightman RM. Comparison of dopamine uptake in the basolateral amygdaloid nucleus, caudate-putamen, and nucleus accumbens of the rat. J Neurochem. 1995;64:2581–2589. [PubMed: 7760038]
9.
Cragg SJ, Hille CJ, Greenfield SA. Dopamine release and uptake dynamics within nonhuman primate striatum in vitro. J Neurosci. 2000;20:8209–8217. [PMC free article: PMC6772736] [PubMed: 11050144]
10.
Schmitz Y, Lee CJ, Schmauss C, Gonon F, Sulzer D. Amphetamine distorts stimulation-dependent dopamine overflow: effects on D2 autoreceptors, transporters, and synaptic vesicle stores. J Neurosci. 2001;21:5916–5924. [PMC free article: PMC6763160] [PubMed: 11487614]
11.
Robinson DL, Heien MLAV, Wightman RM. Frequency of dopamine concentration transients increases in dorsal and ventral striatum of male rats during introduction of conspecifics. J Neurosci. 2002;22:10477–10486. [PMC free article: PMC6758730] [PubMed: 12451147]
12.
Borland LM, Shi G, Yang H, Michael AC. Voltammetric study of extracellular dopamine near microdialysis probes acutely implanted in the striatum of the anesthetized rat. J Neurosci Methods. 2005;146:149–158. [PubMed: 15975664]
13.
Kissinger PT, Hart JB, Adams RN. Voltammetry in brain tissue: a new neurophysiological measurement. Brain Res. 1973;55:209–213. [PubMed: 4145914]
14.
Adams RN. In-vivo electrochemical measurements in the CNS. Prog Neurobiol. 1990;35:297–311. [PubMed: 1980746]
15.
Wightman RM, Jankowski JA, Kennedy RT, Kawagoe KT, Schroeder TJ, Leszczyszyn DJ, Near JA, Diliberto EJ, Viveros OH. Temporally-resolved catecholamine spikes correspond to single vesicle release from individual chromaffin cells. Proc Natl Acad Sci USA. 1991;88:10754–10758. [PMC free article: PMC53009] [PubMed: 1961743]
16.
Thévenot D, Toth K, Durst R, Wilson G. Electrochemical biosensors: recommended definitions and classification. Pure Appl Chem. 1999;71:2333–2348.
17.
O’Neill RD. Microvoltammetric techniques and sensors for monitoring neurochemical dynamics in-vivo: a review. Analyst. 1994;119:767–779. [PubMed: 8067534]
18.
Pantano P, Kuhr WG. Enzyme-modified microelectrodes for in-vivo neurochemical experiments. Electroanalysis. 1995;7:405–416.
19.
Wilson GS, Gifford R. Biosensors for real-time in vivo measurements. Biosens Bioelectron. 2004;20:2388–2403. [PubMed: 15854814]
20.
Fee M, Leonardo A. Miniature motorized microdrive and commutator system for chronic recording in small animals. J Neurosci Methods. 2001;112:83–94. [PubMed: 11716944]
21.
Hardin G. The tragedy of the commons. Science. 1968;162:1243–1248. [PubMed: 5699198]
22.
Antoniol G, Tonella P. EEG data compression techniques. IEEE Trans Biomed Eng. 1997;2:105–114. [PubMed: 9214790]
Copyright © 2007, Taylor & Francis Group, LLC.
Bookshelf ID: NBK2571PMID: 21204385

Views

  • PubReader
  • Print View
  • Cite this Page

Other titles in this collection

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...