Implantable myoelectric sensors eliminate many of the drawbacks of skin-surface sensors. The question is, are they worth invasive surgery?
The Need for Better Myoelectric Sensors
Most myoelectric sensors currently used to control bionic limbs are built into the prosthetic socket so that they press against the skin where they can detect the movement of certain muscles.
The problem with this approach is that residual limbs often change size and shape, causing the sensors to shift position relative to their target muscles. Different arm positions can have the same effect, while excessive sweat or dry skin can interfere with myoelectric signals.
Pattern recognition systems are one possible solution to this problem. They use more sensors, resulting in more data, and also allow for quick recalibrations to match changing circumstances. But even they are affected by many of the same skin-surface issues.
As far as we can tell, the first serious work on implantable myoelectric sensors began in 2003.
The concept was simple: eliminate the myoelectric skin-surface issues by implanting the sensors directly into the target muscles.
The implementation was not so simple. Suitable sensors needed to be designed. They needed to be powered wirelessly and to transmit their data the same way. And a new control system had to be developed to process data received simultaneously from many more sensors and at different depths in the muscle tissue.
The control system is particularly important. To grasp an object with a bionic arm/hand using just two myoelectric sensors on the skin’s surface, an above-the-elbow amputee might have to perform the following tasks:
- switch control to the elbow;
- move the elbow;
- switch control to the wrist;
- move the wrist;
- switch control to the bionic hand;
- select a grip pattern for the bionic hand;
- close the hand around the object.
To make matters worse, the user has to repeat the same specific muscle contractions to perform each task, which can be tiring.
The ability to read data from eight myoelectric sensors and to detect deeper muscle signals makes it possible to identify and interpret each of these intended movements simultaneously. This means that the user can simply attempt to move his bionic arm as he would a natural arm and let the control system figure out how to make it all happen.
That is not only a huge time-saver for the user but also much less exhausting!
The First FDA-Approved Clinical Trial
The first FDA-approved clinical trial of implantable sensors took place in 2014 involving a system called IMES (Implantable Myoelectric Sensor System), developed by the Alfred Mann Foundation. Eight small sensor cylinders were implanted in a patient’s arm as follows:
The results were encouraging. The test subject reported an increased ability to intuitively control the IMES System compared to previous myoelectric and other prosthetic devices. Just as importantly, he reported an increased desire to wear his prosthesis.
This was not a short trial. It lasted three years. To get a sense of where it ended up, check out this video from 2:15 onward:
Ossur’s Lower-Limb Initiative Using IMES
Shortly after the preceding trial began, Ossur announced that they had successfully used IMES in a lower-leg experiment with two amputees wearing a Proprio Foot.
Later, in 2019, they put out the following video showing the same concept expanded to include a bionic knee:
Unfortunately, in November 2019, the Alfred Mann Foundation sold or licensed the IMES technology to Ossur. This is unfortunate because it means that we have likely lost the ability to track IMES as a separate technology and we’ll have to monitor its evolution via Ossur products instead.
Other Implantable Myoelectric Sensors
IMES is not the only use of implantable myoelectric sensors. We’ve also seen them show up as part of various projects involving Targeted Muscle Reinnervation (TMR), and as part of Integrum’s e-Opra Implant System, which combines osseointegration with an integrated neural interface and the embedded myoelectric sensors:
However, we have not been able to track the progress of implantable myoelectric sensors in general. They appear now to be more of an optional part of other solutions rather than an independent solution.
Implantable Sensors Versus Improved Skin Surface Sensors
Implantable sensors seem like a big technological improvement but they come with a serious drawback: surgery.
Not only is surgery risky, painful, and expensive, it also isn’t very scalable. Even if implantable myoelectrical sensors turn out to be an essential part of bionic limb use in the future, we can’t easily perform that surgery on the tens of millions of people around the world who need it.
Meanwhile, the quality of skin-surface sensors has been improving, as have the pattern recognition systems responsible for interpreting their data.
We’re also starting to see sensors embedded in flexible materials like electronic skin:
With this kind of flexibility and adherence, not to mention the greatly increased density of sensors that can be embedded in electronic skin, it is not difficult to imagine myoelectric sensors moving from the socket to some kind of stretchable liner in the not-to-distant future.
If these surface systems finally reach the point where they can equal the detection abilities of implantable sensors, then it’s likely that widespread use of the latter will cease.
How Amputees Should Use This Information
One of the reasons that we devote so many articles to surgical techniques that can improve the use of bionic limbs and the systems used to control those limbs is because we know that amputees are trying to make wise decisions about their futures.
Surgical techniques like the Agonist-antagonist Myoneural Interface (AMI), Targeted Muscle Reinnervation (TMR), and Regenerative Peripheral Nerve Interface (RPNI) are easier to recommend because they are not tied to specific bionic devices and they also have other benefits such as pain reduction.
Implantable myoelectric sensors do not fall into this category. Indeed, as with any technology that may eventually have non-surgical alternatives, their ongoing use is not guaranteed. So, if it were us, we’d consider this type of solution only if it were packaged as part of a complete system such as Integrum’s e-Opra Implant System or Ossur’s product combinations with IMES.
For a quick summary of all related surgical techniques, see Surgical Techniques That Improve the Use of Bionic Limbs.
For a complete description of all current upper-limb technologies, devices, and research, see A Complete Guide to Bionic Arms & Hands.
For a complete description of all current lower-limb technologies, devices, and research, see A Complete Guide to Bionic Legs & Feet.