Medical Electronics Forum

Medical Electronics Forum

Medical electronics bring a variety of special challenges for the electronics designer. Not only must medical devices meet stringent safety and reliability requirements, industry trends like consumerization and miniaturization, and the need for many products to be disposable, sterilizable, or nonmagnetic, place additional demands on the design engineer.

Here, representatives from a variety of manufacturers involved in the industry discuss these issues and their impact on the design and development process.

Designing for the common man
Electronic Products: More and more medical devices are being sold to and used by end-users�regular consumers�as opposed to medical professionals, creating new challenges for design engineers working in this area. What are your perspectives on this ongoing trend of “consumerization”?

Scott Pavlik (Product Manager, Precision Signal Processing Group, Analog Devices): Demographic factors such as healthcare costs and the aging population are creating product opportunities for companies that make blood pressure monitors and thermometers to develop other patient-monitoring types of products going forward. From an IC manufacturer standpoint, the challenge is not only to meet cost pressures associated with these products, but also to meet specific performance, regulatory, and safety requirements that are not typical in other markets that we serve.

Scott Pavlik, Precision Signal Processing Group, Analog
Devices

The clinical products have their own set of requirements, particularly from the regulatory agencies. Some consumer products have these requirements some do not. Safety is always a requirement.

I'd say the biggest design challenge is power, because a lot of the emerging medical products are based on either portability or small size.

In the case of patient monitoring equipment, for example, there are fairly stringent specifications near DC, so a low-noise amplifier is needed. The challenge is to provide the necessary performance on low power.

Paul Magill (Chief Technology Officer, Avo Photonics): The focus for products that we're building is moving the testing from the hands of the doctor or the hospital and into the hands of the patient on a regular basis. The general focus for our customers is “how do I take that testing, give it to the customer, and allow them to do it in a way that is reliable and that is safe?”

Paul Magill, Avo Photonics

Our customers want to move this type of testing from what I would call invasive testing where you're actually making some kind of contact measurements�drawing blood, taking a tissue sample, or something like that�to a noninvasive kind of scanning technique that can be handheld, can be portable but, number one, becomes a noncontact testing method�using optical scanning technique to determine material levels in the blood, for instance. This removes a large part of what could be a liability�when the end-user is not an experienced, educated, trained, technical healthcare professional.

Timm West (Applications Engineer, ETA Circuit Breakers): We provide supplemental protection for the medical devices, so our concerns are mainly safety, price obviously, fitting the application�for example, single-pole, two-pole, what have you�but having all the necessary approvals.

Timm West, ETA Circuit Breakers

On a personal level, I am one of the consumers. My daughter has asthma and we have a nebulizer at home which she does daily. And sure enough on there there's a little rocker switch that is nothing more than a switch and a circuit protection.

Javad Mokhbery (President and CEO, Futek Advanced Sensor Technology): We do see the challenges with regard to the commercialization of the medical products. The challenge that I see is making sure that the right technology�I mean in a way, a partnership�is required to provide the type of product that would be available in the market because the demand is there.

Javad Mokhbery, Futek Advanced Sensor Technology

It is the partnership of the right companies together. The technology exists, but it is a matter of how the project is handled or how the product is designed�the combining of field partners who can make it available and make it possible. We have seen that if the product is easy to use, it has been received very well.

Keith Chipman (Product Manager, IRC Wire and Film Technologies Division): IRC has participated in medical electronics for many years. The resistors are used primarily in stressful environments, where a designer needs to handle high-voltage applications, high-power applications, or high-pulse-energy type of applications.

Keith Chipman, IRC Wire and Film Technologies Division

And as we move towards consumerization of many of these medical devices, some of the same requirements get pushed down into those devices�for instance the increasing demand for the portable or home-use defibrillators has driven us to work quite closely with a number of these manufacturers toward miniaturization and cost reduction while maintaining the quality and reliability of our resistors.

It is important as design becomes smaller and smaller to work as early as possible with the providers of the passive components to adequately specify and provide an opportunity to improve those products if necessary to support the miniaturization.

Robin Tichy (Technical Marketing Manager, Micro Power Electronics): I'd like to thank Scott from Analog Devices for the great lead-in when he said that power is one of the biggest challenges. We definitely see that from the battery pack manufacturer point of view.

Robin Tichy, Micro Power Electronics

One of the challenges is the expectation from the consumer point of view that a medical device is going to have all of the functionality of their iPod.

So it's going to have a color screen and all kinds of bells and whistles and wireless communication, and all of these things are very power hungry, and it becomes a serious challenge to design a battery that delivers more power but is both lighter weight and smaller size.

On another point�the cost reduction and still maintaining quality for a product like a defibrillator, which is really going to save someone's life�there are some unexpected challenges for a battery pack.

We manufacture about 70% of the automatic defibrillator batteries, and one less obvious challenge there is the shelf life of the battery. In terms of consumer acceptance, the defibrillator may only be used once, or hopefully is never used. But if it is, you want it to work, even it's been sitting on a wall for years, and batteries don't have an infinite shelf life. It's a real challenge.

Bob Procsal (Applications Engineer, OPTEK Technologies): We work with lots of different medical equipment manufacturers, and they want to bring the cost down of the products that they have and enhance reliability. They're also looking at an RoHS-type product, even though it's not required for the majority of the medical industry because the cost of a non-RoHS product is getting more and more expensive as RoHS is staying down in price.

Bob Procsal, OPTEK Technologies

The types of things that we've been working with companies on are home kidney dialysis machines for analyzing different blood or serums, home IV systems, and food pumps. They're also looking for different wavelengths for good skin promotion, such as acne, collagen enhancements�all those kind of things.

They want to do tanning booths where they don't have the bad wavelengths�they just have the tanning wavelengths.

We've talked about the killing of the germs by using the ultraviolet and the 265-nm range. One of the main concerns is the wavelengths of the devices that they need to be utilizing, and what's the tolerance that they could have on those wavelengths to meet the requirements.

Doug Rasor (Vice-President, Worldwide Strategic Marketing, Texas Instruments): We think the growth in this kind of consumer or personal medical devices is really attractive. The consumer part of it we think is the largest segment.

Doug Rasor, Texas Instruments

A lot of the basic needs that are driving that are the spiraling increase in healthcare costs, plus just quality of life�if you have the ability to do more of that at home yourself, there's a lot of peace of mind and finding problems earlier and so forth, and all those are great benefits.

Certain categories like defibrillators obviously are Class 3 FDA devices and have very unique requirements. But in our experience, in a lot of the other devices like glucose meters, digital thermometers, blood pressure monitors, and so forth, a lot of our customers can use a lot of our commercial products. There's some additional documentation we sometimes have to provide, but we can provide a lot of components we sell in other segments.

And the “care abouts” that we continually hear are the ones that I think we talked about here: it's got to have a small footprint on the board�so fewer components are better�low power, and of course low cost. And I don't know how you would order those but it's probably roughly in that order.

So, we find ourselves doing more systems on a chip, and an example could be for in a glucose meter. Really common among all the manufacturers are things like a decent-performance microcontroller that runs at ultra-low power, some data converters, LCD drivers, battery management devices�and so we've integrated those onto a single chip specifically for glucose meters.

Noel Giamello (Senior Director, Systems Solutions Business Unit, Sharp Microelectronics of the Americas): From Sharp's point of view, we supply a great deal of the LCD technology to the medical industry, both in the hospital and for consumer use. And in the mobile LCD space, there are quite a few characteristics that are really important for medical applications that we've run across�for example, the viewing characteristics, how clear the display can be under various light conditions. This might be text or an image, it might be in a defibrillator or something as simple as a small medical PDA that's used for blood analysis in the field, or some of these are creeping into the consumer space now too where people can purchase these for home use.

Noel Giamello, Sharp Microelectronics of the Americas

And there are several different types of display technologies that are out there whether it's reflective or transmissive, or there's a term called transflective that's kind of the best of both worlds thrown in there where the LCD manufacturer�and we do this as well�can tailor the reflectivity or the transmissivity of the LCD to accommodate the power requirements of the portable device and also the type of environment that it might be used in.

So if you're in an ambulance in varying light conditions, if the screen is too reflective, you really can't see the data that's on there and it could be time critical, just depending on what's going on with the person.

The other thing that's important for medical applications is that there are FDA requirements involved and whenever you change a component or a technology within the device, they have to go through a new FDA qualification which can be rather costly for the manufacturer. So longevity of the product is extremely important and the company supplying the device needs to ensure a length of time that the device will be available off the shelf, as well as a good mean time between failure so that the device functionality is there when needed.

The cost of consumer medical devices is going to be more sensitive than what's being sold in the hospital environment, and some of these devices are being done offshore for cost considerations. At the same time, they still need to go through FDA and it needs to be easy for the engineer-designer to integrate the panel with the MCU, and with the LCD controller that's present in the application.

Robin Tichy: One of the great ways of keeping cost down is to use components that are common in some other kind of device. In the specific case of batteries, many medical device manufacturers can decrease the cost of lithium ion batteries by using the form factor that's used in laptop computers, which is essentially a cylinder that's 18 millimeters in diameter and 65 millimeters long.

Noel Giamello: Are there any moves towards using fuel cells?

Robin Tichy: Fuel cells are definitely coming onboard, but it's been a slow process for them. I think really what's going to be one of the first things to become available is a product where the fuel cell can charge the battery pack in an emergency or in a remote location, thereby increasing the runtime.

Paul Magill: I'd just like to talk a little bit more in general about the reliability requirements that medical products may have. Super reliable parts are by their very nature, going to be more expensive, and I think the industry needs to decide�as you get to a consumer-oriented level and you're selling products that may have a one-, three-, or five-year lifetime, and not a 30-year lifetime�do we really need to build in and test-in the kind of reliability that AT&T used to have on telephone pole equipment that had to be out there unattended for 25 to 30 years?

The other thing too that I think really needs to be addressed with the industry�and I see this in other industries�is regarding the tests that people frequently do, for instance, the accelerated test for reliability. Frequently these do not have field data associated with them anymore, which means the stress data may have no actual relationship to the reliability of the product.

What kind of reliability do we want for these various products, and is the testing�the accelerated life-testing that we're doing�appropriate to say that we have a 10-year lifetime or a 15-year lifetime or a 20-year lifetime? And I think that this is a big question.

Size, cost, or power?
Scott Pavlik: One comment on miniaturization is markets such as handsets and PDAs, where miniaturization has been an ongoing process, has provided the technology for smaller, lower cost solutions to enable medical consumerization. Integrated circuits are becoming increasingly available in smaller form factors such as chip scale packages or direct attachment of die, for example.

And in addition to that, fine-line geometry CMOS processes are pushing down the supply voltages, power and chip area, getting more functionality in a given area of silicon. So, technology trends that have pushed miniaturization in handsets can be utilized in the medical field.

Doug Rasor: I think the other markets have really provided a lot of the impetus for all of us in the chip world anyway to invest in chip-size packaging and stacked die and all sorts of crazy things that a decade ago we never would have imagined we'd be doing at the cost points that we're currently at.

From a technology standpoint you assume that your ability to handle the power and the size is really a design decision or a matter of will. Put the right engineers on it and these are problems we can solve. The biggest challenge we find is can you solve them at a cost that meets the goals of customer.

Noel Giamello: Sometimes the cost requirements might be somewhat unrealistic also, because many times these technologies are developed not primarily with medical in mind, but might be some fallout from let's say the automotive world, or the handheld world, where you need to meet an extreme cost performance target, and size, and you need to do this in millions upon millions of pieces so that reliability becomes extremely high.

Robin Tichy: It's amusing that power is always brought up as one of the top three concerns, but the battery pack manufacturers and designers tend to be brought in relatively late in the design process. So from our point of view, hitting a cost target is a challenge, but there's also a real challenge in integrating the battery pack into a medical device that's essentially already designed.

Doug Rasor: One of the challenges we have, if it's GE medical calling, or Philips medical, or Siemens, or some of the big folks, it's easy to get teams of people across the system.

Our challenge is that we find that a lot of the most innovative devices get designed at small companies who ultimately may even get acquired by these big guys. So if you're a 30-person start up and you've got some IP from the university and you're off designing something, your source of information and interaction is probably with manufacturers' reps or distributors or maybe just via the Web.

Timm West: I too would like to agree with Robin. She's doing battery packs, we're doing supplemental circuit protection–it seems like we are the last consideration.

Bob Procsal: Whenever we talk to the medical industry, the main thing that they really consider their main objective is to have the best performance. Reliability comes in as number two, and then cost comes in as number three.

Because of the new concepts and the new products that are available, cost can be brought down in lots of different ways. We provide some products that are replacing some assemblies in equipment at major medical companies that cost them $100 to build the way they were built it in the past, with an item that cost about $10 from us.

Part selection prescription
Electronic Products: Many medical products have special issues�such as the need to be disposable or sterilizable, or submergible or nonmagnetic�that can present challenges for designers. What's your perspective on how these issues can affect the design and part selection process?

Doug Rasor: Where we tend to get specific requirements back on us�and in the case of especially Class 3 medical devices�we really have to design specific test structures into the chip to allow complete testing from outside the chip and to guarantee zero (DDPM) quality levels, very much like in the automotive industry and it really affects design and the thing we have to do internally to be able to supply the customers.

Noel Giamello: From the part selection of the design process we try to get as close to the engineer as possible right away.

First, you want to get a feel for where the company is going in terms of their product line and what their mission is, their strategy. They're trying to be the highest quality medical device, is it just something that they want to proliferate across the board? Is it sold off the shelf without a prescription, or is it through a prescription�that affects the cost of the device and the type of design constraints that are usually placed upon the manufacturer.

The cost of these medical devices can go up considerably because of the FDA Class 2 and Class 3 requirements that they have to go through, and the engineering requirements that are placed upon the team.

If power can't be mitigated very easily, we'll try to get some strategies on how to handle that issue. Maybe we'll use an LED backlight instead of a cold cathode tube to power the display, maybe the MCU needs to have a superior sleep strategy to save power.

But again, we'll try to work with them as closely on the drawing board as possible.

Bob Procsal: When we talk to the companies that are doing these designs it is important that we both have an understanding, which Noel was just talking about, that we have a good specification of what they're looking for.

If we look at the autoclavable situation, most of that is a pretty hot environment and you really can't put it in an autoclave, so you have to come up with some other way for sterilization. The UV at 265 nanometer seems to be one of main ways to do that without getting the heat in there and without destroying the plastics or other characteristics of items that you have.

Referencing the disposable market, you have home IV systems and food pumps and other products where they want to make sure that they don't have bubbles in the system and have a fluid going through the tube�optoelectronics is one way of ensuring this inexpensively and very accurately.

Robin Tichy : One of the other requirements that we see for batteries that are used in medical devices is knowledge of the operating time to empty. Often the fuel gauging is a big differentiator, especially for life support or Class 3 medical devices.

So companies that manufacture ICs are offering very, very accurate field gauges, and they're very popular with the medical community because you can get a very accurate countdown to when the battery is going to lose power.

The autoclave presents a huge problem for the battery pack. Batteries generally cannot be exposed to such high temperature and in some cases batteries can reach a state which is called thermal runaway, which can present not only damage to the battery but actually a hazardous situation for anyone in the room.

Keith Chipman: As far as design and part selection, I think it's important for the designers to work with component manufacturers early in their design and evaluation, and sample and test the parts that would enable them to have a second round of evaluation if they're evaluating tradeoffs in terms of size, power, and voltage selection.

It's important for them to get it right because in systems that do require FDA approval, once that's done and they're pretty much locked in, it's much more difficult to change designs at a later date.

For disposable medical products, it is important that the component manufacturers use materials that can be disposed of in a safe manner. RoHS-compliant parts are a great first step.

For nonmagnetic applications we've had some interesting calls for resistors that would be used in the cable pods for EKG or EEG where they want to provide monitoring during an MRI image scan. We do have units that don't use steel caps�they use a plating for the electrical contact that's nonmagnetic.

Javad Mokhbery: Actually we have had products in the medical industry which do meet autoclavable, nonmagnetic, or submergible requirements. However the challenge is when these requirements are given with the additional requirement of miniaturization, as well as lower cost.

So one of the approaches that we had taken in the past such as with the autoclavable requirement, we teamed up with the medical instrumentation designer to see if we could come up with a noninvasive rather than invasive approach that would require the sensor to be autoclavable every time they use it.

So my point is that lots of times some of these requirements really require good teamwork and a partnership to understand exactly what the purpose of the application is.

Timm West: I'd like to reconfirm what basically everyone else has said so far�that we work closely with the OEMs to better understand the needs of the device.

And whether that includes UL1077 approvals, VDE, CSA, what have you, possibly offering a system solution, maybe a water splash protection is the best way to go for this certain application or an actuator guard to prevent inadvertent operations, or possible signalization to let you know when something is occurring or tripping.

Paul Magill: Since we're primarily focused on all optical or optoelectronic components, we're not really dealing with the need to make disposable products since that already deals with the issue of sterilization, because in making an optical measurement you're not making a contact type of measurement.

However, for solutions for those customers that have autoclavable or submergible requirements, we borrow heavily from our experience in telecom and military product developments where these requirements have been in place for a number of years.

The kind of hermetic solutions that you find in the telecom industry, we're applying towards medical products that we're developing for customers today.

Scott Pavlik: I confirm all that has been said about the importance of working closely with customers and understanding the issues and their needs.

To make another point similar to the one about size reduction of components and how the handset industry drove that, I think meeting RoHS standards for IC manufacturers such as Analog Devices, Texas Instruments, and others, have placed us ahead of the curve for disposable products.

Small is big
Bob Procsal: One of the things that's happening in the optoelectronics industry is we're providing surface-mount-type products which allows us to miniaturize some of the standard optoelectronic products that are available today.

We should talk about heat for miniaturization, and surface mount. By using custom designs you can take all these concepts, put them together, and come up with an optimum solution for the customer, both in size as well as utilization of the product to meet the system requirements.

Javad Mokhbery: With regard to the miniaturization, I guess we need to look at the reference. What does it really mean? Every year we go the MDNM medical show and we present very small size sensors, which we believe we have challenged ourselves�10 years ago, it was 1-in. diameter; right now, it is a quarter inch diameter. But every time we go there, we are asked, “Can you make it smaller?” So it is all based on what the requirement or what the references are.

Paul Magill: Everyone wants to miniaturize. But we see in some instances a move to instead of just, for instance, having a light pipe into the body, with all of the actual sensing equipment remoted, we see now a move toward getting the sensors down into the body and then the information coming back out. So a focus on more localized direct measurement as opposed to what is now done mostly in a remote sense.

Bob Procsal: I think in summary we can all say that the product is never small enough, it never has enough performance, reliability is never good enough, and it always costs way too much.

[Group laughter]

Noel Giamello: One of the issues that comes up is not just the size. A customer may specify the size of the panel and they'll usually do that with some market studies, like the Apple iPod�the 2.46-in. LCD display that's extremely popular now. There are many customers who want to use that size LCD everywhere, no matter what the application seems to be.

But the size of the bezel that's around the LCD can be an issue, because that can affect the overall size of the product or handheld device that it might be going into.

And the other thing that customers seem to get a lot of angst over is the thickness of the LCD module, even though the engineers would like to have a really simple interface to the LCD module. These two demands compete with each other.

And as many of you probably know, the LCD product requires varying voltages and timings, so we try to simplify that by putting as much of the intelligence as possible on the back of the glass. This can affect the thickness and the size in a good way.

Robin Tichy: For batteries also, there's a drive that even if we have to work with a certain volume, in some cases, it's advantageous to move to a thinner product.

So using the Motorola Razr as an example, the battery that's used in that is actually a lithium polymer battery, where essentially the casing for the battery is something like a coffee bag material and that allows it to be very thin.

But I also had a question for Noel, if he finds that the resolution of the screen is a big challenge for them from the display point of view?

Noel Giamello: Yes, actually the push is for us more from the cell phone side because that's where most of the volume is�that and the mobile handheld area.

In the medical arena, they need to see a lot of data in the smallest area as possible, again, because of power constraints and the size of the handheld device, but the text and image needs to be very clear. And some of the text can be very tight.

If this is imaging in a hospital, they get very stringent in terms of how many pixels might be missing on the screen and whether the image is rendering as true as it would on a CRT. We get held to that extreme.

And in that case the resolutions of the monitors are 1 Meg by 1 Meg for some of the MRI and CAT scan imaging. So as you go down in the miniature side of things�that handheld device�it's at least quarter VGA now.

Keith Chipman: One comment on miniaturization, it's important to understand the physics of a lot of devices, in my case especially resistors.

The industry has been successful at miniaturizing resistors to where they're almost microscopic, and that works fine when you're working with low voltages and signal-level type of circuitry.

But where there's a need to dissipate from a fraction of a watt to several watts, the issue is to conduct the heat away from the resistive element so you usually end up with a physically larger device.

Same thing with the ability to absorb large amounts of pulse energy, and I'll go back again to the defibrillator example. There are times that we've got resistors that are required to absorb anywhere from 25 joules to 200 joules of energy.

Timm West: I think someone had mentioned earlier of doing more with less. And this is an idea that's not only designated to the medical device industry but all the other industries that we service.

So what E-T-A has done is to offer a wide variety of circuit breakers and a wide variety of circuit breaker technologies in many varieties to handle this miniaturization issue that's come up from across the board.

Doug Rasor: What we're being asked to do by our cell phone clients and some consumer device manufacturers in terms of smaller size packages and integrating multiple functions on fewer pieces of silicon really play into this miniaturization thing in medical. And so we do see reusing a lot of those techniques and even some of the specific devices in some of these medical devices.

Scott Pavlik: Doug is absolutely right. As was mentioned earlier about never small enough, never low enough in power, and never high enough in performance, sometimes customers don't understand the challenges associated with mixed-signal technology and creating integrated solutions.

For example, at first glance one process technology that gives you low power may not provide the performance needed in a front-end design, while reducing size. Close collaboration with the end customer to figure out the right solution is usually required.