Design Challenges of Portable Medical Monitoring Systems
Today’s portable medical devices produce design challenges, both in the form factor of the device and the electronics inside.
By Jonathan Bearfield, Texas Instruments
click to enlarge
Figure 1. Portable medical monitoring systems. |
Government and medical organizations are driving to put patients more in control of their health care today. To allow patients more time in the home versus hospitals or
doctors offices, the medical industry is leveraging portable, remotely linked medical monitoring systems. These include everything from blood glucose meters to portable electrocardiogram (ECG) systems.
Challenges with "portable" medical electronics are the need for increased portability with remote connectivity, while maintaining the quality and responsiveness to any data collected. The term "portable" used to mean the equipment had wheels and could fit through a door. But today, we see things differently. Now many medical devices are fully transportable, some even potentially "wearable." This, of course, brings with it design challenges, not only in the form factor of the device — but also in the electronics inside.
If we tear apart any portable medical monitoring device, we find that they have five basic building blocks (see Figure 2):
•Display and display interface
•Battery and power management
•Biometric sensor interface
•Data interface
•System microcontroller or digital signal processor (DSP)
Obviously the specific performance of each block varies from system to system.
Displays
Whether informing the patient of their temperature or electrocardiogram (ECG) readings, the visibility of the display is an important performance attribute, which is partially driven by having the right backlighting solution. Whether leveraging an inductive or simpler charge-pump-based topology, as portable systems are powered by batteries, a backlighting solution with a wide input voltage range reduces the need for additional regulation in the system. Selecting boost or buck/boost solutions that can easily operate from multiple battery chemistries provides greater flexibility in the design. Naturally, solution size and overall power efficiency are key system careabouts, so leveraging devices with a high level of integration and advanced packaging becomes more efficient and cost effective.
Touch screen control (TSC) is a key driver in making portable electronics easy to use, but it also allows the equipment to become significantly smaller since traditional keypads can be removed. TSC allows menu driven function selections, fine tuning of input and output data displays, and keeps the “keys” large and easily usable. When implementing TSC, one important aspect to review is the electrostatic discharge (ESD) handling capability of the solution selected. If the TSC circuit can not dissipate the energy from an ESD strike, the energy passes through to, and damages, the central microcontroller/DSP. Other TSC considerations are resolution as compared to screen size, conversion type and speed, and overall power consumption.
Sensor Interface and Signal Chain
Whether reading temperature, pulse, blood glucose or other biometric sensors, implementing the appropriate signal chain is paramount. An instrumentation amplifier is the first stage of the signal chain
click to enlarge
Figure 2. Major functional blocks in portable medical electronics. |
(see Figure 2), like an INA326, a micro-power amplifier with low input offset, low drift and great DC accuracy with AC performance. In most implementations you are trying to find a microvolt level signal within milivolts of noise. Since the target signal has AC characteristics, an amplifier that works well with a high-pass filtering scheme is needed. Leveraging an auto-zero or auto-cal feature can further simplify system compensation requirements.
Typically, the second stage is a low power operational amplifier, like an OPA376, which has wide bandwidth, rail-to-rail input and output (RRIO), and offers an excellent level of precision. Features like zero-crossover produce signals with linear offsets over the entire input common mode range. This means that the microcontroller does not need to run additional algorithms to correct for shifts and offsets.
Transitioning into a good delta-sigma or successive approximation analog-to-digital converter (ADC) is the next stage in the signal chain. Features like single cycle filter settling, and convert-on-command simplifies the design requirements around the ADC. It also improves conversion speeds and allows larger source impedances. In multi-channel systems, features like global synchronization provide coherent signal acquisition, allowing multipoint signal sources to be compared within the same clocking cycles.
With appropriate layout and component selection you feed a clean, accurate and truly meaningful signal into the system microcontroller/DSP.
Microcontroller/DSP
Medical monitoring devices produce endless amounts of raw data. Storing the data and processing trends, noting changes, providing feedback, supporting the ability to link up to larger systems, and performing diagnostic algorithms are often critical functions of the system controller.
Balancing the systems processing needs against the power consumption limitations is a critical decision. Targeting DSP-level data processing but only allowing for the power budget of a low-power microcontroller, like the MSP430, would cause a conflict in the design. However, by leveraging newer DSP technologies, and power supply topologies, and implementing several power levels and
standby modes can help a system get sports car level performance at economy car level gas mileage. This means that some of the processing bandwidth may need to go towards managing power consumption. However, while an MSP430-type controller manages the system in standby, sleep and wake up transitions, a DSP drives overall system performance to help create a system with the best of both worlds. If DSP is active only when processing needs to be done, average system power consumption remains low, with the peaks happening only when the DSP is awake. DSP power surges can be supported by implementing super-caps or other energy storage devices to minimize brown-outs and improve system runtime. Given the performance and integration levels of the latest microcontrollers, sophisticated applications needing real-time processing at very low power becomes possible with devices like the MSP430FG461x.
Battery and Power Management
Simple systems may use disposable batteries, because their power draw is low enough to keep the overall cost of replacing batteries low. Larger systems leverage various rechargeable battery chemistries and pack sizes. Features
Important Notice
System and equipment manufacturers and designers are responsible to ensure that their systems (and any TI devices incorporated in their systems) meet all applicable safety, regulatory and system-level performance requirements. All application-related information in this article (including application descriptions, suggested TI devices and other materials) is provided for reference only. TI disclaims all liability for system designs and for any applications assistance provided by TI. Use of TI devices in life support and/or safety applications is entirely at the buyer’s risk, and the buyer agrees to defend, indemnify and hold harmless TI from any and all damages, claims, suits or expense resulting from such use. |
such as dynamic power path management allow the system to draw power independent of the battery charging path. This allows a device with completely discharged batteries to be used as soon as it is plugged in, versus waiting for the batteries to charge. There isn’t always time to wait when the need to use a medical system arises. The ability to track the true impedance of a battery versus simple voltage measurements or coulomb counting is another important feature. Since battery voltage does not drop off linearly, voltage tracking doesn’t give a true reading of battery life, especially since the middle third of the voltage scale comprises 60 to 70% of the discharge cycle time. Coulomb counting does not compensate for battery aging, and without learning the new capacity over time, it "assumes" the state of the battery versus actually knowing it. Impedance tracking allows the system to calculate the remaining run-time to within one percent error over the entire life of the battery, enabling longer run-time by allowing the system to access all of the usable energy in the battery.
As system operation is critical in medical electronics, another important feature is battery authentication. This is a means of validating that the battery in the system meets the original equipment manufacturer’s (OEM) requirements via an encrypted device ID. Using the wrong battery pack can impact system run time, as well as damage the system, or even cause a fire.
In general, making power decisions at the front end of the design cycle helps to define system-level tradeoffs that may be necessary to meet the portability and run-time targets for the equipment.
Data Interface
The medical electronics data interface migrated from cabled RS232, to wired and wireless Ethernet connections and near field and longer range wireless connections. The newer interfaces allow hospitals to network all of the equipment in the building, as well as a patient’s home.
When a patient comes home from the hospital, he could be remotely connected to their doctor via a wireless sensor pack on their body, which links to a monitor connected to their home security system. The entire system ties into the Ethernet or a medical call center to receive around the clock monitoring in the privacy of their home. Wireless interfaces like Bluetooth may have a play here. Also consider Zigbee and other very low power wireless solutions by Chipcon products from Texas Instruments, whose SmartRF® technologies are networking homes as well as industrial settings. Aside from power consumption, data rate and range are the two key careabouts when selecting a wireless interface. A 2.4 GHz solution provides worldwide coverage, a high data rate and duty cycle, with many channels. But lower frequencies increase the signal range. For multi-channel, full body monitoring, the patient probably is confined to their home, if not a bedroom. In this case, the range could be limited, but the data rate maximized. In situations where we are monitoring just a couple of sensors, range may be more important than data rate. At the end of the day, the solution choice must fit within the overall system power budget.
In Closing
Our future holds innovations such as home body scanners where we stand in front of it and see our "doctor" on an LCD display. The virtual doctor may be anywhere in the world while we could be at home, in the office or on vacation. Even today, portable medical devices and monitoring systems are putting medical support exactly when and where you need it. To enable medical device manufacturers to develop these innovative products, we need the right infrastructure outside of the electronics as well as the right semiconductor components inside. For greatest success, semiconductor suppliers will want to consider the features and requirements of portable medical products, define performance specifications for each of them, and understand their space and power budget limitations. Ultimately, this can help manufacturers of medical devices to reduce any rework, and optimize the overall design from the start.
About the Author
Jonathan (Jon) Bearfield is an end-equipment marketing engineer providing complete system solutions for the High Performance Analog Team in Texas Instruments.
Texas Instruments
P.O. Box 660199, MS 8711
Dallas, TX, 75266-0199
© 2008 Advantage Business Media
|
