back to article Robo-surgeons, self-driving cars face similar legal, ethical headaches

As drivers contemplate computer controlled cars, physicians get to ponder self-driven surgery tools. In a research paper published on Wednesday, "Robot Autonomy for Surgery," UC San Diego assistant professor Michael Yip and PhD student Nikhil Das explore the growing role played by surgical robots and the issues raised by …

  1. John Smith 19 Gold badge
    Unhappy

    "lack judgement,.. touch isn't as refined.., they're expensive,.. lack the ability to improvise.

    But what you've built one of you can build hundreds (or 1000s) of.

    Note that fact the laparoscopic robot could be used autonomously and has been around for 2 decades but is not allowed to be.

    Surgeons are a well financed and well organized trade union (amongst other things) and most of them will fight this.

    OTOH hospital administrators will be eyeing up massive (potential) reductions in malpractice insurance, 24/7/365 availability and better than average survival rates (I'm hearing an Austrian voice saying "BONES 1 performs first autonomous operation under supervision on August 3rd 20xx. It continues to operate with a perfect patient safety record for ten years. A decade later all hospitals have them installed. Surgical graduate numbers have dropped by ninety percent as they cannot perform to the standard set by the robot").

    Personally I think the developers of these ideas have gone for grandiose visions which antagonize surgeons. It is an astonishingly complex environment to deal with.

    A better approach IMHO would have been to identify sub tasks the surgeons would be happy to outsource, so they can get onto the next patient. Opening up and closing up come to mind. Assistance with placing,tracking and removing surgical equipment and consumables in the patient also come to mind, allowing a very small team to act like a much larger number of hands and eyes.

    My instinct is it would be better if such robots could use conventional surgical instruments, but possibly not with conventional "hands," to ensure more precise translations and rotations. Then again keyhole surgery already uses specialized instruments the surgeon has limited direct control over.

    I suspect, like human surgeons, surgical bots will specialize. Ideally they will share a common mounting (not necessarily in the floor) with different sizes depending on how heavy the work involved is will be. I can imaging the orthopedic 'bots being quite hefty, to cope with joint replacements, likewise heart operations for chest cracking. OTOH going for other major organs slimmer, lighter appendages would be possible.

    1. Nattrash
      Meh

      Re: "lack judgement,.. touch isn't as refined.., they're expensive,.. lack the ability to improvise.

      I'm working in this field for more than 2 decades now, and agree with many of the arguments you mention. I mean, the advances in automating processes in medicine like for example automated sample preparation (think making your blood, urine, tissue sample ready for diagnosis) have been massive. Same for imaging; the things we can do now with for example MRI, PET, or computer CT are much, much better then when I started in the 90s. For treatment the same applies. The AccuRay radiotherapy system mentioned in the article has helped people fight cancer much better while reducing radiation exposure (= 3D imaging, site localisation, and dynamic spacial radiation pattern). And one could argue that this isn't robots at work. But if you look at them carefully, you'll find there is, although it might not look like the Start Trek holo physician. But...

      Having developed these kind of systems, there's one thing that we shouldn't forget. These are all great tools, and we could extrapolate to surgery, but there are some serious PITAs concerning complete automation, like for example a robot placing a nice new titanium hip. And that is interhuman variation.

      As you mention correctly, robots are great as repeating something tirelessly. But they are rubbish at adaptation, improvisation, correlation. That where that silly, fuzzy human mind of us works so very well. So, yes, you're right - not all surgeons are keen on these surgo robots. But is it really the scare of ending up on the dole? My experience is that the biggest problem is we all as patients are too individual, too unique. And yes, we are much alike (actually the base principle of medicine), but then again we are not.

      Just think about it. Don't jump into surgery immediately, but look a bit closer to home. For many, many years automatic tissue sample preparation and scanning has been worked on. And it kind of works, but (IMHO with good reason) a person still looks at it in the end. Why? We techs all know how difficult, inconclusive, inflexible, prone to false positive, negative (and so on) image analysis can be. Sure it is easy to measure blood or potential cancer cell nucleus size, roundness, colouring, or just how many there are. But what does it mean? This is where tech is rubbish, and we humans are not too bad.

      I came to learn the hard way that medicine is NOT an exact science. And THAT is the biggest problem for automation (like this) in medicine.

      Oh, and people trying to make shitloads of money trying to sell automated medical crap of course... ☺

      1. Alister

        Re: "lack judgement,.. touch isn't as refined.., they're expensive,.. lack the ability to improvise.

        I agree with Nattrash.

        It is not possible to reduce medicine to a set of rigid algorithms, there is just too much variation from the textbook cases.

        Doctors, surgeons and other medical professionals are not just trying to protect their jobs or act like luddites when confronted by this new tech.

        There is a genuine fear that patient outcomes will worsen, or patients will be harmed by either mis-diagnosis, or inappropriate or misguided surgical procedures, because these medical AIs do not take into account the infinite variety of human conditions.

        As someone who was for many years a paramedic, I know there are times when you have to ignore what your training teaches you to expect and just make an intuitive decision based on experience to ascertain what's wrong with a patient.

        There is, I'm sure, an overwhelming body of evidence for patients who do not show symptoms, or present with incorrect symptoms of their illness or injury, unfortunately this evidence is probably mostly anecdotal, and therefore dismissed as irrelevant by those who write the algorithms for medical AIs.

        1. a_yank_lurker

          Re: "lack judgement,.. touch isn't as refined.., they're expensive,.. lack the ability to improvise.

          I agree with the previous posters. One of the major problems with any fully autonomous system are the cases that were never considered in the design. Or the fact that patients will react differently to the same medication and dose. Handling these situations takes discernment and experience plus one can not wait for the software to be written, tested, approved, and uploaded.

          1. Anonymous Coward
            Anonymous Coward

            Re: "lack judgement,.. touch isn't as refined.., they're expensive,.. lack the ability to improvise.

            Err, slow down a bit.

            First off, this is nowhere near an established, stable field of medicine, secondly I only partly agree with "it's too difficult for a computer" - that is, if you look at it from the perspective of REPLACING medical staff which is IMHO the wrong focus because, indeed, humans are not like cars.

            I see these developments mainly from an augmentation perspective - new tools. I've never been a fan of not keeping humans in charge (although in plane technology, Airbus clearly disagrees with me), but I do think that good tools can enable us to do more or be far more precise than we are now.

            Admittedly I'm a non-expert, but I have clocked up enough decades in electronics, mechanics and computers to know we still keep inventing new and wonderful things that make life better. Do we need to be damn careful with how we apply those ideas? Oh yes, certainly when you start in the medical field, but I also see major benefits if we do so sensible.

            I think we're a fair distance away from full autonomy. We can't get cars to work yet either, but even there we're far further than we used to be so I expect progress. As long as we frequently cull the sales people in this field :).

    2. Anonymous Coward
      Anonymous Coward

      "expensive"

      Given what a surgeon in the right specialty makes in the US, a robot that's able to schedule operations back to back, day and night, 365 days a year is worth up to $3 million a year. That robot will get cheaper and better performing over time, while surgeons will only get more expensive (unless/until their salaries start being pushed down by robots)

      In countries where doctors don't make what they do in the US robots will take much longer before they can compete, but it should happen eventually.

  2. tiggity Silver badge

    "What happens when region-locked abortion machines get hacked for unauthorized usage?"

    You raise a toast to women wrestling back some control over their body / life choices and not just being treated as a (rights free) life support system for an embryo.

    .. and auto doc workarounds such as Prometheus film spring to mind

  3. ma1010
    WTF?

    Eh?

    Who pays when the euthanasia bot malfunctions?

    Well, if you were angry at still being alive instead of dead, I suppose you could sue the bot maker? Or just go jump off a bridge?

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like