The MIS2DU12 is the smallest (2.0 mm x 2.0 mm) ST accelerometer suitable for medical applications with a typical current consumption of only 6.1 µA in normal mode with antialiasing and 0.47 µA in ultra-low-power mode, enabling implantable devices like pacemakers to last 10 years or more. Both the mechanical and ASIC parts of the MIS2DU12 have received optimizations that help keep the current down while still enjoying a sampling rate of 1.6 Hz in normal mode. Furthermore, while most of our customers use this sensor for medical applications, we can also envision it in industrial settings that require years of field operation on a tiny battery.
Underestimating the impact of accelerometers on low-power systems
It’s easy to forget
Working on a machine like a pacemaker has numerous challenges, but one engineering constraint that rarely gets much attention is the impact of electronic devices on power consumption. For instance, in a recent article by Sriramoju, Iyengar, and Srivathsan from the Mayo Clinic1, the physicians offer a comprehensive and fascinating list of complications that can arise from pacemaker use, including “generator failure”. And while they do briefly address the risk of a battery running out of charge, they don’t explain why or how an electronic component running amok could be at the center of this issue. Indeed, for many, this is simply an engineering problem that can feel almost divorced from how the device performs medically.
It’s important to remember
However, there’s a case to be made for the fact that the power consumption of a pacemaker is intricately linked to its performance as a medical device. Because pacemakers use a tiny lithium battery, every microamp counts. Consequently, given the extremely tight power budget, something as small as an accelerometer can drain the battery enough to negatively impact the product itself and create serious issues for the patient. Indeed, such a sensor is crucial, as the information it continuously gathers helps doctors adjust therapy based on a person’s activity or posture. That means both its ultra-low power consumption and typical current consumption noticeably impact battery life.
It’s critical to get it right
Moreover, having an accelerometer with ultra-low power consumption is not enough on its own, because the sensor must still be precise, especially in a noisy environment like the human body. Unlike the idyllic conditions of a laboratory, an accelerometer planted inside a person will need to filter out muscle tremors, environmental shocks (such as a heavy object dropped on a floor), vibrations reverberating inside the person, and signals from other parts of the body that could trigger a false positive. And a poor signal‑to‑noise ratio could have catastrophic effects, as the pacemaker’s analysis of the patient’s condition could lead to inappropriate therapy adjustments, potentially endangering the subject.
MIS2DU12: Optimizing performance and power consumption
Why use an anti-aliasing filter?
Avid readers of the blog will know that the vast majority of sensors have two domains: an analog one responsible for the sensing element and passive components, among other things, and a digital one, which hosts the device’s application-specific integrated circuit (ASIC). During normal operation, the sensing element captures information as an analog signal, which passes through an analog-to-digital converter before being processed by the digital circuit, which outputs the results. To filter out high-frequency noise that may contaminate the information sampled by the sensor, we apply an anti-aliasing filter, which improves accuracy and reduces the amount of post-processing required by the host MCU, among other things.
Why move the anti-aliasing filter?
The MIS2DU12 is unique because it addresses power-consumption and accuracy challenges simultaneously by moving its anti-aliasing filtering into the analog domain rather than keeping it in its digital domain, as is commonly done today. Traditionally, a sensor will have an actual output data rate that is much higher than requested by the application. For instance, a 100 Hz ODR will actually take significantly more readings than 100 Hz suggests to improve the signal-to-noise ratio. However, moving the filter to the analog domain means the accelerometer can use a much lower actual ODR, since there’s much less signal-to-noise correction required. And reducing the actual ODR directly lowers power consumption.
The idea of putting the anti-aliasing filter into the analog domain is far from new, but the industry has largely moved away from it in recent years because it made the overall sensor larger and more expensive to produce. However, thanks to its manufacturing expertise, ST can keep prices reasonable while ensuring that the MIS2DU12 remains the smallest accelerometer, despite its low power consumption and its ability to withstand temperatures (+85 °C) compatible with the human body. As a result, we are seeing early design-in with customers working on medical applications.
Why wait to try the MIS2DU12?
The new accelerometer offers output data rates ranging from 1.6 Hz to 800 Hz and can sense acceleration at ±2 g, ±4 g, ±8 g, and ±16 g. It also includes a self-test feature that can detect free-fall, wake-up, and single- or double-taps. While many customers in the medical field tend to use a custom PCB for their proof-of-concept, we still offer the STEVAL-MKI255A adapter, which houses an MIS2DU12 and is compatible with the STEVAL-MKI109D evaluation board. Engineers can just use this kit and run MEMS Studio to start designing a proof-of-concept in minutes.
-
Anil Sriramoju, Shruti Krishna Iyengar, Komandoor Srivathsan. Navigating Complications in Cardiac Pacemakers: A Comprehensive Review and Management Strategies. Rev. Cardiovasc. Med. 2024, 25(8), 299. https://doi.org/10.31083/j.rcm2508299 ↩︎
