Biomedical MEMS do more than just "sense"

Most engineers think of MEMS devices primarily as sensors. This makes sense, because that's where the bulk of the market is, as airbag triggers in cars, pressure sensors, accelerometers, or gyroscopes. But there are many other attractive applications for MEMS technology, as I learned in a conversation with Dr. Jeffrey T. Borenstein, Distinguished Member of the Technical Staff and Director of the Biomedical Engineering Center at Draper Laboratory (Cambridge, Mass), while at the TechConnect World Conference & Expo in Boston.

For biomedical applications, Draper and others are investigating MEMS technology as a critical enabling technology for many non-sensor situations, including large-device fabrication as well as tiny, sophisticated actuators.

For example, in one project, a silicon wafer is micro-etched with tiny channels to act as a fabrication master. This master is then used to make many "copies", which are then stacked up as layers. The result is an artificial organ to be used as a supplement or even a replacement for the liver, as the blood flows through the many channels. A similar device and situation would clearly much better for kidney-dialysis patients, who now must go to a clinic for blood cleaning three times a week, typically.

Critical note for those who assume that "smaller is better" when it comes to process geometries, and think that dimensions in the tens of nanometers are the needed--similar to those of today's digital ICs--keep this in mind: Dr. Borenstein said that the appropriate biomedical-device features are on the order of ten microns, which is three orders of magnitude larger than our state-of-the-art ICs, since blood cells are around 5 microns in diameter.

There are also interesting developments under way for applications in precision, internal medicine delivery using MEMS-based actuators. Presently, a drug must be delivered either by injection, or orally, and thus often causes unavoidable collateral damage to other parts of the body besides the target area. Also, the patient may have problems adhering to the delivery schedule and protocol. Even so, the medicine may not reach the right spot, in the right dose, or with the right timing.

For example, there's some indication that it may be possible, with the right medicine to spur regeneration of the frequency-tuned hair cells in the inner ear which respond to sound (vibration) and are critical to converting the incoming sound energy to nerve signals. Dr. Borenstein said they are exploring doing this via a MEMS-based, microfluidic reciprocating push-pull pump which periodically squirts the dose into the inner-ear area. The thinking is that such a burst-type injection is preferable to a slow, continuous application, since it is less intrusive and thus less prone to rejection, where the bodies proteins attack the injection as a foreign material.

Don’t expect to see this system too soon-- problems are not just physiological. While there are options for powering such a device, including batteries and wireless recharging, the refilling of the implanted drug container is another challenge. Draper is working with Massachusetts Eye and Ear Infirmary, a leading institution, to test the concepts, and while they have preliminary test data on animals, it will be several years before will have similar data on human trials. That's certainly a big difference from your conventional new-product cycle for most electronics we design.

(For additional information, see here, here, and here.)