In the world of medical science, mobile circuits are quickly evolving into less cumbersome devices. They are providing more information and using less power. At the same time, they're improving quality of life. Although this evolution is only beginning, it's already resulting in smaller process capabilities, longer-lasting batteries, and advancements in medical understanding and circuit designs. Currently, electronic devices monitor blood glucose levels, blood pressure, and many other medical conditions or concerns (FIG. 1). The future isn't just in monitoring medical conditions, however. It's in controlling them.
Thanks to circuitry advancements, many medical devices are becoming mobile, fully implantable, and unnoticeable to the user and the outside world. Clearly, these developments will eventually lead to long-term implantable circuits. Before these devices can be born, however, many challenges and tradeoffs need to be resolved.
First, one must determine the specific needs that will be satisfied by the device. To control most medical conditions, it's vital to have timely information. Real-time data enables a device to quickly adapt to a diet, activity level, or medication. A situation can then be controlled before it becomes critical or life threatening. To provide this kind of information, however, a device must be mobile, easily accessible, and capable of constant monitoring. It also has to be self-adaptable and extremely reliable.
Many of these requirements sound easily achievable. It's only when they're combined that the trouble starts. For instance, envision a device that's mobile and continuously monitoring. Clearly, battery life will be a consideration. Now, add easy accessibility to the picture. Obviously, the device will have to be small in size. Luckily, this combination supports mobility. But it contradicts battery life and memory requirements for self-adaptation. Now, factor in that it must be very accurate and cannot be altered by external forces (heat, light, water, sweat, being bumped or other movements, etc.). The implications of protecting the device simply become impractical.
Think about how the human body works, how it adapts to changes, and how the internal organs are protected. The only conclusion is that the ideal device must be wireless and implantable. Of course, it will put severe constraints on battery life, size, adaptability, and data-transfer capabilities. Even more importantly, a medical procedure will be required to install such a device.
Yet a wireless implantable medical device offers one major advantage: the potential to improve a person's quality of life. Ideally, an implantable device should place minimal restrictions on the physical activities in which a person can participate. The goal is to enable the individual to monitor/control a medical condition without worrying or feeling restricted by the medical device.
For many individuals, such devices could provide real-time information combined with corrections to a given medical circumstance. Subsequently, the medical condition could become a non-issue in everyday activities. For this vision to become a reality, however, a number of medical and design issues must be overcome.
Initially, the success of a wireless implantable device will be based on the installation procedure. It must be low risk. Due to the limitations of sensor or battery life, it also should be able to be done multiple times. These requirements put a great deal of pressure on the medical, electrical, and sensor communities. They must determine how to obtain critical medical information from the body. In addition, these communities need to figure out how to control a medical condition without sacrificing vital information or placing too much risk on the patient during installation.
Sensor and circuit designers face severe limitations when trying to control a device's size. Size plays an important role in determining how and where a medical device can be installed. It also factors into the device's overall cost. Among other cost variables are the medical risk with installation and the device's lifetime. For example, being smaller may make a device easier to install. But a smaller device will usually cost more while limiting battery or device life.
Next to installation, the success of wireless-implantable devices rests on adaptability. These devices must change with the medical condition or the body. Over time, the way that the sensing device interacts with the body will be modified. In addition, the device must adapt to activities like exercise, being physically bumped, or shifting within the body.
Generally, controlling a medical condition requires different algorithms or coefficients for every individual. It may even require additional/other information from the body. If the devices aren't designed with adjustability or adaptability in mind, the advantages of wireless two-way communication can be lost.
The final factor in the success of wireless implantable devices is the length of time that they can remain in the body. "The longer the better" is a good rule of thumb here. Initially, the body will require time to adjust to the implant. The size of the device and the invasiveness of the installation procedure are just some of the factors that determine the body's adjustment time. If the battery or sensor life is too short, the actual sensing time—which takes place after the body has adjusted—may be limited.
For a long-term wireless implantable device, power consumption is generally the most critical requirement. It determines how long the device can be in the body. It also can decide the size of the battery. In many cases, the battery is the largest part of a long-term implantable.
To meet the individual needs of many different medical conditions, wireless devices must be long-term, small, and fully implantable. Long-term implies fewer visits to the doctor and less medical bills. The device's small size translates into an easy installation procedure and shorter adjustment time. Lastly, a fully implantable device should provide the necessary measurement modeling from a controlled protective environment in the body. At the same time, it offers the freedom to pursue activities without the worry of removing or damaging the sensing equipment.
Wireless technology is the thread that makes this combination possible. It provides the capability for adaptability and flexibility in controlling algorithms and sensors. At the same time, it allows a device to be fully implanted. When confronted by an implantable medical device's stringent requirements, however, wireless designers only find themselves in control of certain tradeoffs. The critical tradeoffs that control the success of the implant involve size, accuracy, adaptability, and power consumption.
Size versus accuracy
In integrated circuits (ICs), high accuracy usually requires larger transistors. Typically, the requirements for accuracy are based on transistor matching. Because transistor matching is dependent on area, the best way to get good matching is to increase transistor sizes. Better transistor matching also can help with voltage offsets and gain control. Consequently, good accuracy demands good matching, which in turn requires larger transistors. These transistors impact the size of the IC or die.
The die size also can be increased by the accuracy and complexity of the algorithms. If the accuracy requirements become too stringent, the design complexity rises. The size of the digital or analog circuitry will then grow, affecting the overall die size. Even though size and accuracy are critical issues, some of the algorithms and accuracy calculations can be done externally with the help of wireless communications. The ability to do these calculations comes at the cost of power consumption. Nevertheless, external control is another degree of freedom to work with when balancing the tradeoffs between size and accuracy.
Size versus adaptability
Here, the goal is to minimize the things that one would like to adapt. Yet the designer must still be able to get as much information and control of the implantable device as possible. Two-way wireless communication makes programmability possible. To allow adaptability without huge impacts on die size, however, smart algorithms and register-controlled state machines are needed. Adaptability usually requires memory. Regardless of whether that memory is used to store data or algorithm coefficients, it still requires space. The more adaptable the device is, the more memory or space it will usually require.
A lot of thought also will need to be given to the area of control and adaptability. Simple and smart sensing techniques need to address changing algorithms and state machines. Such devices are usually the easiest to adapt in a system.
Accuracy versus power consumption
Ultimately, power consumption is the element that will limit how accurately an implantable device can be monitored and controlled. Say a device is being monitored and controlled continuously. The number of transmissions or receptions per minute, hour, day, or year can quickly consume a battery—especially when combined with the power that's required to complete a task. Consider both the number of bits and the number of times that the device transmits and receives. The number of bits transmitted relates to both accuracy and power. More bits imply more accurate results. But the device will consume more power during a transmission.
Another area to consider regarding accuracy and power is the distance at which the device must transmit. Transmitting over a large distance will consume more power. The designer must take into account the accuracy of the transmission, bit error rates (BERs), error detection, and correction algorithms. In the tradeoffs between accuracy and power, important roles are played by the number of samples taken per day and the number of bits that must be stored or sent. Also, factor in whether a sample needs to be fixed if it's corrupted or resent if it's missed.
COMMON DESIGN STRENGTHS
If a medical device is more adaptable, the information that it obtains tends to be more accurate or "useful." But accuracy in controlling a medical condition isn't just a function of obtaining data. It also affects the understanding of the factors that help to control the condition.
An adaptable device should be able to provide more accurate information by doing the following: reading additional information from the sensor, averaging or filtering the results, measuring another sensor, or adjusting the algorithm according to other factors. In the latter case, the factors might include the time of day, level of physical activity, or types of food eaten. Through the proper use of adaptability, it's possible to obtain accurate knowledge. Such knowledge is vital to controlling a variety of conditions in a changing body.
Note that size and power consumption are directly correlated. It's the size of the battery that usually determines the amount of available power. The total time that an implant can be in the body is based on the battery's power capacity. To regulate a device's size, it's crucial to control its power. Because the battery is a major factor in the device's overall size, it also affects the adjustment time and the ease of the implant procedure. Consequently, low power is generally the most important issue for the designer to address.
The implantable device has no choice but to be small. It must ensure a simple installation procedure, short adjustment time, and the ability to receive protection from the body. To meet the area requirements of a small device, however, a number of parts will need to be adjusted. This requirement will affect circuit boards, die aspect ratios, and battery shapes.
Obviously, the external components have to be minimized. In addition, the total part count must be redeemed, such that some components will have to be placed on the integrated circuit itself. The sensors will have to be designed to conform to or around a battery shape. The resulting package should allow for easy installation.
To enable wireless communication, antennas need to be incorporated on the circuit board. This requirement poses another potential problem. Generally, small antennas are very inefficient. They also are known to reduce output power. Such issues must be eliminated before an antenna finds a home in an implantable medical device. Remember that power consumption and battery size are the major factors in keeping device size under control.
When the time comes to choose circuitry, however, the designer must face a new pressure: the critical nature of precision or accuracy in diagnosis. To determine the type of electronic circuitry needed, one must understand the sensor's capabilities and how accurately it can measure. Note that if simple analog and digital techniques are used to obtain the data, they will keep the measurements in their truest form. They also will provide the most accurate results to the user. Determining the number of bits for analog-to-digital converters and digital-to-analog converters should be straightforward. When trying to obtain reliable results, however, consider the offsets, mismatches, noise, and gain requirements on analog circuits.
To determine the accuracy required by the system, one must know how the sensor's dynamic output range changes over time. If the sensor's output signal becomes smaller over time or with lower battery voltage levels, accuracy will be compromised. Power consumption may even need to be increased. Remember that sensors with small signal levels should not be sampled during a wireless transmission. The noise from the power amplifiers could distort the sensors' signal or the sampling circuitry.
Adaptability can be divided into two areas. First, there's the adaptability of the part as it automatically tries to monitor and control a person's condition. The second area concerns how the doctor or patient sends wireless information. They could be sending this information to change how the sensor reads and transmits data. Or they could be altering the way that the algorithm or state machine performs a task. When talking about either form of adaptability, the designer must walk a fine line. Otherwise, size may become an issue—especially for the die and sensor. To understand the level of adaptability that needs to be incorporated into the design, figure out what critical information is needed. Then, determine the information that can be obtained from a sensor.
In the development of automated adaptability within the device, smart simple algorithms make the circuit adapt to changes in the body. Note that when designers try to develop adaptive systems, they usually require more power and space. To minimize that tendency, smart simple algorithms need to be developed. Most smart algorithms use the previously obtained data or history that's stored in the device. This data can help the algorithms determine or adapt the coefficients automatically.
Smarter algorithms also can help to predict when serious conditions may occur. Such predictions are based on previous condition trends. Because the information is sent wirelessly from the body, medical science can learn the different correlations between the controlling factors. This information also can be used to build better smart algorithms.
The concept of sending the information from the implant to obtain better control leads to another area of adaptability. With a wireless adaptable system, information can be sent to a receiver outside the body and used in various ways. The two-way wireless communication allows the flexibility to externally control internal algorithms.
An external device can make controlling decisions while storing history. At the same time, it can alter how the sensor functions or what information is received. This external controlling capability can be utilized to save power in the implant. At the same time, it provides powerful algorithms and history-building information that can give real-time feedback to the implant.
This form of adaptability may be the most powerful. Because the external systems aren't constrained by power, their batteries can easily be changed. With the power limitation reduced, more complex continuous controlling algorithms can be tested. The testing won't reduce the device's internal life. This form of wireless capability allows the external system to be placed within a reasonable distance from the user. Most importantly, the user still has his or her freedom when it comes to physical activity.
The complexity of controlling the medical condition is really what determines how to make the system adaptable. The single-sensing devices with well-known controlling factors may opt for a single enclosed system with internal control. More complex sensing systems, which deal with compounded medical conditions, may require an externally adaptable system. These complexities must be determined in the beginning of the project. The designer can then choose the best way to give the implant longer life. Plus, he or she can retain the ability to change the monitoring or controlling mechanisms for better results.
During every phase of the design cycle, one must consider power consumption for implantable devices. When dealing with this type of device, sensor life and power consumption dictate how long the device will reliably function in the body. For this reason, power constraints need to be taken into account at the very beginning of the specification phase. They are key to defining and determining the implantable's different features.
To manage any low-power system, different power-down procedures will need to be examined. Because a wireless system usually dissipates a lot of power during transmit and receive modes, unique power-saving ideas also need to be explored. Planning a simple low-power solution isn't easy. Attention must be focused on the full system power-down procedures as well as the wireless-transmit and receive power-saving modes.
To ensure successful power-down procedures, it's critical to know the different circuitry blocks' sequence of operation. During the beginning design phases, learn how often the sensor needs to be read. Consider what circuitry is necessary for the device to retrieve the desired information. If the sensor is required to be on at all times, for example, what is the minimal circuitry that's needed to retrieve this data? Can the remaining circuitry be powered down until it's needed? Say the sensor can be duty cycled. Can it be quickly stabilized to allow the retrieval of data and then power down for some time period?
Both cases require a low-power internal oscillator. That oscillator will need to be functional in order to keep the timing when power is provided to the different circuitry blocks. If it's possible to use the duty-cycle method, the power savings can be substantial—especially if some of the blocks are only turned on less than 10% of the time.
The best form of power savings is when the whole circuit—aside from the timing oscillator—can be powered down and then turned on for a short time. Consider the following example. A device is "on" 1 ms and then "off" for 9 ms. The resulting power savings is 90%, which means the designer is getting nine times more battery life out of the device. A major part of power savings is discovering how to power down certain blocks and then get them powered up and stable before reading the sensor. Be sure to make a complete analysis, as a block will sometimes burn a lot more power on startup than in normal operation.
As mentioned, trying to save power may affect the accuracy of the sensor measurements. With a wireless system, this issue can become even more challenging. To save power while the device is transmitting information, the designer must consider when as well as how much information needs to be sent. Depending on the desired distance, a major portion of the current that's being consumed will be in the output power amplifier.
To conserve power, consider limiting the number of transmissions and bits that are being sent. If the number of bits being sent is large, different compression algorithms can be utilized. For another way to save power, power down the device's transmit section and store the data over a number of readings before sending it. This approach may not appear to save much power, as each bit takes the same amount of power to be sent. But the startup and stabilization sequence of a power amplifier can burn a large amount of power. Power savings also can be affected by reducing the number of times that the power amplifier is powered down and then back up again.
During a wireless device's receive mode, the average power consumption can be even worse than it is during transmit mode. Most wireless systems are in receive mode whenever they're not transmitting. Typically, the receive circuitry is lower in power than the transmit circuitry. Generally, however, the receiver section must be on for longer periods of time.
Clearly, this method does not serve as the best way to conserve power. If the design requires the receive mode to be on continuously, current consumption can be reduced by only using the receiver's received-signal-strength-indicator (RSSI) circuitry subsection. The receiver will then use the amplitude of the RSSI circuit to determine if the full receiver should be powered up. For this step to occur, the preamble must be long enough for the receive block to wake up without missing any of the data. Because the transmit signal will be coming from the outside world, the preamble of the data can be fairly long. Thus, the implantable only needs the RSSI circuitry to be powered up most of the time.
For another power-saving approach, utilize a long preamble in a power-down duty cycle. The receiver will then be able to be turned on and off. This mode of operation requires a very quick starting receiver. The act of allowing the receiver to sleep and then duty cycle is known as Sniff Mode (FIG. 2).
As seen in Figure 3, Sniff Mode is a series of patent-pending ideas. First, it allows the receiver to sleep for a period of time. Then it very quickly wakes up and does a "sniff" to determine if a desired signal is present. If the receiver detects a strong enough signal, the whole receive chain wakes up and begins receiving data. When Sniff Mode duty cycles the receiver, power savings is achieved for on time versus off time.
With mobile wireless systems, every mechanism that can save power must be considered. This statement is especially true for implants. Advancements in wireless communications have opened the door for devices to be implanted. Now, newfound creativity is being funneled into the development of medical devices.
This article has examined different ways of addressing size, accuracy, adaptation, and power consumption when designing a wireless implant. Ultimately, each medical device will have to work through the different needs and tradeoffs of the medical condition that it will serve. Only then can it provide a unique solution. The resolution of device issues is very important. Yet it doesn't even come close to the overall goal of the medical device, which is to improve a person's quality of life.