Wireless medical implants represent an exciting $17 billion market worldwide that includes cardiac pacemakers, cochlear implants and neurostimulators. The sector is projected to grow at more than 22 percent annually through 2009 — with some segments such as neurostimulators growing at nearly twice that rate. Wireless-enabling medical products provide a tremendous advantage to both patients and caregivers — including greater patient ease of movement, continuous data feeds, greater quality and reliability of data reporting (compared to patient self-reported data) and ultimately improved patient outcomes. However, when you do this for devices implanted within the human body you take such advantages to the next level.
As a result, design engineers are more frequently being asked to explore the adaptation of existing medical product designs for in-body wireless use. Yet designing for the often harsh, signal-blocking environment within the body is no small task. Implantable devices contain local digital processing within the unit that can be optimized and adjusted for the specific individual. A wireless connection built into an implant provides the low-level control necessary for downloading previously logged performance data and updating device control parameters through a low-level messaging protocol.
When enhancing medical devices with any kind of wireless connectivity, it is important to remember the link quality and data integrity are vital for FDA-approval if it is used for diagnostics or therapeutics. For medical implants in particular, power consumption is also a major problem, leading to highly optimized solutions for ultra-low power consumption from the battery.
Delicate Balancing Act
From the design engineer's perspective, this represents opposing challenges when designing wireless implants. The product must consume miniscule amounts of power from the battery, while providing a high-quality and reliable wireless link for the underlying data. The design engineer has to find an optimum balance among power consumption, range and data rate within the radio regulatory constraints.
Professional wireless products, such as cellular telephones, have the luxury of space and power to employ the full gamut of methods to protect the underlying data and manage interruptions on the radio link. For example, the low-power ZigBee telemetry standard supports resilient data signaling and is well-matched for external medical devices in and around a hospital. It can setup and manage its own “ad-hoc“ mesh network over the air, for communication between devices such as bedside care monitors or wireless control of foot switches in an operating theatre. However, although power efficient, ZigBee does not achieve the ultra-low power consumption needed for implantable devices.
The alternative design course is to keep things simple by relying on commercial off-the-shelf (COTS) integrated RF devices. These usually protect data by sending repeated messages and using simple check codes, which is neither a power efficient nor a reliable method for sending data within a reasonable response time.
Getting it right depends on striking a good balance between the application needs and the strength of each wireless approach.
Power Consumption Challenge
Wireless links for pacemaker/defibrillator implants or neurostimulators typically provide a low-bandwidth channel for small and infrequent control messages and a higher bandwidth reverse channel for downloading stored data records (which can be extensive) from the device. This approach in the radio transceiver matches the RF channel to the quantity of data required in both directions, resulting in an asymmetric radio link.
Although this imposes a difficult constraint on the radio designer, the use of an asymmetric link is essential in maintaining low power consumption from the internal battery. A high-bandwidth link can transmit a given quantity of data in a shorter time, thereby saving battery energy. In fact, it is the simple energy equation (Energy = power * time) that constrains and governs many of the radio design considerations for wireless medical implants.
Whether powered by primary or rechargeable batteries, the challenge for wireless implants is to find an ultra-low power RF solution for transferring data. Design for such systems is not trivial and requires specialist knowledge of radio link budgets to understand how far the transmit power can be dropped before completely losing the signal inside the body.
This effect is compounded by the physical dimensions of the implant, which are often quite small. Unfortunately, “size matters” for an efficient antenna design and the electrical length (physical length/operating wavelength) determines the amount of radio signal that can be launched from a given transmitter. For a given area, the antenna efficiency improves at higher frequencies, but this must be balanced against the increased attenuation in the flesh. Too much signal attenuation causes radio dropouts and reduces the operating range of the link.
There is of course a natural “sweet spot” for the operating frequency in wireless implants and studies have shown this to lie somewhere between 200 MHz to 1 GHz depending on the device size and implant depth. In recent years this has driven adoption of the MICS frequency band (402-405 MHz) for human wireless implants in many countries across the world (U.S., Europe, Japan), and the 868/915 MHz unlicensed ISM bands for non-implanted devices.
Most wireless implants are intended for short-range communications across only a few feet, so the transmit power is typically low and close to 100 uW. Although lower transmit powers are possible, the higher data rates are still needed on the reverse channel to reduce transmit time and hence save energy. By combining digital processing functions with the analog radio circuit, further energy savings can be achieved by compressing the data records at source, prior to transmission.
Mixed-signal radio ICs based on CMOS or BiCMOS technology can achieve such ultra-low power solutions by combining the radio and digital processing functions into a single chip.
Data Quality Challenge
Radio links in electronic devices are inherently intermittent and yet it is still possible to transfer high-quality data across such links. Even when using wired connections, medical devices have to protect against data loss and potential breaks in the system. The key medical requirements of data quality and data integrity are met by applying extra data protection mechanisms to the wireless connection. These protection mechanisms are often overlooked in simple COTS designs, because they add a level of engineering functionality that can be ignored in less demanding markets.
It is important to understand and quantify the data loss policy required for a wireless implant, before the wireless system design can start. For example, how much diagnostic data can be lost during an acquisition period without affecting the prognosis? Should gaps in the data be detected? What response time is required from the device? These are all questions that relate to delivery of the data and the underlying quality of service (QoS).
The wireless designer addresses these QoS issues by introducing layered protection mechanisms on top of the underlying data. The main techniques include:
Forward error protection — where the source data is post-coded with supplementary parity data to enable the receiver to correct over-the-air corruption of data bits. For moderately noisy channels, near-perfect correction can be achieved by using high-performance coding schemes. For example, we have designed a power-efficient form of low-density parity check (LDPC) coding that achieves performances close to the theoretical Shannon limit.
Block error detection — where blocks of source data are protected with a CRC code, used by the receiver to detect residual data errors or breakdown of the link.
Automatic message repeat request (ARQ) — where a high-level handshaking scheme is added to verify transfer of whole control messages at both ends of the link, thereby ensuring high data integrity. Each of these mechanisms needs to be balanced with the supporting radio protocol, to find the best combination of wireless link resilience and adequate response time for the device under control.
In today's crowded spectrum the radio designer has to anticipate and manage interference from multiple sources in order to improve the radio link reliability, while simultaneously achieving low power consumption.
By combining analog and digital technology into mixed-signal RFIC design it is now possible to achieve both reliable data integrity and low power consumption in the same device — the essential elements for successful wireless medical implants.