Question: What is the future of robotics in medicine?

Susan Cartwright answered on 21 Jun 2015:

There are lots of possible applications for robotics in medicine. Some are already happening – for example, it is possible for a surgeon to carry out an operation in another city, by controlling robot arms to manipulate the surgical instruments, and using a camera and a headset to see what she is doing – a bit like playing a video game, but in real life. Artificial arms and legs are also improving by the adoption of robotics technology – I expect that in the fairly near future, if you are unfortunate enough to need an artificial hand, you will be able to open and close the fingers just as if it were your real hand, using actuators that can detect the electrical impulses in your nerves. It will probably be harder to give you a realistic sense of touch, but we are already nearly there with the motion.

These are examples of robotic technology remotely controlled by a human being – the surgeon in the one case, and the owner of the artificial limb in the other. Is it possible to imagine “real”, self-controlled, robots conducting operations? Well, it’s possible to imagine it, but I think it will be difficult to accomplish. While computers and robots find easy some things that are extremely difficult for us, like doing complicated mathematical calculations, some things that are easy for us – because we have evolved to do them well – are very difficult for machines. These include making sense of visual information: a camera can produce an image of a scene with much higher resolution than the human eye, but it is very hard for a computer analysing that image to do things that are easy for us, like pick out someone we know in a crowd, even though he’s a long way off and we can’t see his face properly, or read unfamiliar handwriting. That sort of thing is exactly what a surgeon has to do: if she is removing an appendix, for example, she has to be able to pick out the inflamed appendix from the rest of the intestine, tie it off so as not to leave a hole (which would be very dangerous for the patient – letting the contents of the intestine out into the body cavity is a sure recipe for infection), and then remove the appendix itself. Everyone’s insides are a little bit different in how the intestine folds up on itself, so this is not a question of navigating to a particular point – you have to interpret and recognise what you are seeing. This is why it takes so long to train to be a surgeon, and I think it’s quite beyond our current technological level with robot vision. So I think unmanned operating theatres where robots do all the work are a good way into the future!

One possible future direction would be nanotechnology, which aims to develop machines at the molecular level of size. This would open up the possibility of actually injecting tiny machines into your body to treat certain diseases. I can imagine, for example, a machine that would travel through your bloodstream removing fatty deposits from your arteries to reduce the risk of a heart attack.

There are some areas of medicine where it isn’t clear which line of attack will be more successful, robotics or biology. I am pretty sure that treatment of spinal injuries will become much more effective in the foreseeable future, but I don’t know whether the damaged spinal cord will be repaired by using stem cell therapy to regrow the nerves, or by using robotics technology to transmit the nerve signals across the break.

Artificial intelligence, as opposed to robotics proper, might well have a future in diagnostics. It is difficult for a GP to diagnose unusual diseases, because she does not see cases very often and it may have been a long time since she left medical school. I can definitely imagine computerised GP’s surgeries where you or your doctor enter your symptoms and the results of any tests into a computer program, and it comes up with the likely diagnosis and the next steps to take. This might speed up referral of urgent cases to appropriate specialists and improve outcomes.

In short, I think robotics could become really useful in medicine. Of course, any new techniques will have to be thoroughly tested before they are licensed for use on real patients, but there are real opportunities to use this technology to help people.