It’s way past time that everything that ever goes wrong inside our bodies should be fixed by tiny little robots. Way past time. There are still a lot of things that we have to figure out, though, like how to make robots that are small enough and remain controllable and practically useful.
One technique is to leverage the magnetic fields inside a clinical MRI machine to control swarms of small, “dumb” robots, using clever algorithms to move multiple robots in different directions even when you’re restricted to giving them all the same input signals. Plus, you can see on the MRI images where your robots are going and what they’re doing in realtime. Driving them around the body like this is pretty cool, and the next step is getting them to do stuff, such as clearing blockages or delivering drugs.
Research presented last month at the IEEE International Conference on Robotics and Automation (ICRA) by Aaron Becker, Ouajdi Felfoul, and Pierre E. Dupont from University of Houston and Boston Children’s Hospital shows how a small swarm of robots can turn themselves into a Gauss gun, firing projectiles that can penetrate tissue.
A Gauss gun, like a coilgun or railgun, is a kind of a magnetic accelerator, but it works a bit differently. It uses the principle of conservation of momentum to transfer force through a series of magnets, turning stored magnetic potential energy into kinetic energy. You can build one yourself out of magnets and ball bearings like the researchers did:
In principle, the self-assembling Gauss gun works the same way, using little magnetic robots steered by an MRI machine.
Operation of a Gauss gun. (a) Standard design for use outside an MRI scanner shown before and after triggering. Magnetized spheres are red and green. Non-magnetized spheres are gray. (b) Design for use inside an MRI shown before and after triggering. All spheres are magnetized when inside scanner. Image: Aaron T. Becker/University of Houston
Each millirobot is navigated to the target location in sequence, and once there, they attach to each other to form the gun. A special trigger millirobot is sent in last, and when it contacts the rest of the gun, it causes the gun to fire. Then, you use the MRI to navigate everything back out again.
The robots haven’t been “milli-ized” yet, so these first tests aren’t at bloodstream scale, but as a proof-of-concept system, it seems very effective:
The magnets used here are steel, which (inside of an MRI) become more magnetic than neodymium. The projectile is a needle that could be loaded with drugs, or just used to punch through (say) a blood clot. Once these robots are made small enough, they could navigate through the fluids in our body (including the bloodstream and the fluid in our spine) to a target area to perform tasks like “puncturing a membrane to release trapped fluid, opening a blocked passageway or delivering a drug to a tissue location several centimeters from a fluid-filled space,” according to the researchers. It’s not at all difficult to imagine how an injection of robots could travel directly to the site of a problem, treat that problem directly, and then travel right back out again, resulting in a much more effective treatment with fewer side effects for a whole range of conditions.
The next step here is to optimize the entire system for clinical use (which means making it much, much smaller, of course), and implimenting closed-loop control of the robots for effective navigation. We’re not sure how long it’s going to take to get all of this working well enough to start helping people, but you’re looking at the very beginning of some incredibly important techniques in medical robotics.
“Toward Tissue Penetration by MRI-Powered Millirobots Using a Self-Assembled Gauss Gun,” by Aaron T. Becker, Ouajdi Felfoul, and Pierre E. Dupont from the University of Houston and Boston Children’s Hospital, was presented at ICRA 2015 in Seattle.