Last week I visited with a few members of the engineering staff at Abiomed, a medical-devices company in Danvers, MA. Abiomed specializes in cardiac-assist devices, mechanical pumps that help a damaged heart to heal by offloading some of its pumping for a period of time. But their most exciting development is the AbioCor, a fully-implantable (i.e., no tubes or wires passing through the skin) mechanical artificial heart. Some of you trekies out there might remember the episode of Star Trek: The Next Generation (terrible name for the series, which also often featured absurd plots, bad writing, and acting to match) where Captain Picard had to have his own artificial heart replaced. That future isn’t quite here, but it is close: fourteen people have already had the AbioCor implanted in clinical trials. It’s pretty amazing to visit the room at Abiomed where long-term testing of the AbioCor is ongoing – tank after tank with these amazing devices just clicking away.
But, being a wireless type, not a mechanical or medical engineer, I wanted to speak with the Abiomed engineers about the radio used to communicate between the AbioCor and external monitoring and control equipment. I wasn’t really surprised to find that the current radio is very, very simple – 19.2 Kbps, RS-232, 915 MHz. narrowband unlicensed, simple on/off keying (data represented by amplitude only, with equally simple coding to balance the ons and offs), and low transmit power limiting range to only a couple of meters. No network protocols, security, interference management, or anything fancy at all – just enough for basic telemetry (five waveforms, such as motor speed and pressure transducers, at 100 samples/second, along with 84 parameters, such as voltage, waveform key features, battery data, etc., sampled every two beats, and 64 potential alarms) and control.
The heart itself is powered by an internal battery, of course, but this must be regularly recharged by external battery packs connected via what Abiomed calls “Transcutaneous Energy Transfer”, or TET. TET also includes a backup wireless telemetry link. They are following developments in Witricity, a wireless-electricity approach pioneered at MIT, but this technology isn’t quite ready for prime time yet.
So, why wasn’t I surprised by the relatively primitive nature of the radio involved? Two reasons: first of all, the mission of the internal radio is quite simple and nothing fancy is really needed here. Even more importantly, however, a new radio would require rigorous and time-consuming review by the FDA, so new wireless technologies and medical devices don’t come together overnight.
Regardless, a question – would a standardized radio like Wi-Fi make sense here? The argument against is that communication in this case need only take place between the implanted device and the vest or similar garment holding the batteries for the TET. It might be better, then, to integrate Wi-Fi and perhaps even a cellular radio into the external electronics, using these as relay points to off-site diagnostic, monitoring, and control facilities. Security and integrity will be vital in this application, which might in fact need to scale, eventually, to millions of units. But while the external-relay concept is a more likely approach, it could also be argued that implanted Wi-Fi would provide a more robust signal, improved security and integrity, more bandwidth in the event such becomes desirable, and better range – covering, for example, a patient’s entire residence when the external TET connection isn’t available (when bathing, for example).
Or maybe a radio conforming to the FCC’s already-approved Medical Implant Communications Service (MICS) would be the way to go. This is also, however, functional over only a very limited range – but assuming the relay strategy noted above, that’s all that’s required. It’s always a question of the best tool for the job – and, countering my earlier thought on this matter, this might be one instance where Wi-Fi does not, in fact, take over the world.
