Accept for a second that robot surgeons exist. Who will check they're up to the job – and how? Medical robots should be split into different categories of autonomy, just as driverless cars are, a group of academics proposed Wednesday in Science Robotics. As should be obvious, surgery requires extensive training. In the US at least, students have to successfully get through university, postgraduate medical school, and …
Once they are cutting you open, you are in $DEITY's hands, no matter if it's a human or a machine. What frightens me most in this article is the mention of "a care assistant robot at level four". Good thing I'm old enough to die before this happens. The current underpaid and over-managed care slaves are bad enough.
I'm not sure I agree with you there. If a Robot can do the (unenviable) tasks of emptying the bed pans, and rolling Mrs Lardass on to her other side, it would free up the human care assistants to actually focus on the high points of their jobs, talking to the patients, helping their mental health and working on mobility. Things like that can actually have a huge effect, but too often with Care Services cut to the bone, the nurses have to focus on the other repetitive tasks and have no paid hours to do the good stuff.
Admittedly, my Scenario relies on funding remianing in place to hire the human care workers alongside the robots and not the companies and governments cutting the money even tighter, so its just the robots in work...
Call me a pessimist but your last paragraph points out exactly what will happen. Here in the US is all about profit. The medics of old are no longer going the extra mile unless they get paid. The corporate masters rule all. Sad that medicine isn't the "noble" profession it once was.
To Play devils advocate - whose Definition of "better than" do we use? The Insurance Companies definition - where the Robot surgeon is cheaper than the human surgeon? The Hospitals definition - where faster is more important, so more patients can go through? The Surgeons opinion - impossible! A Robot can never be better than a surgeon! (based on my one experience of drinking with a surgeon at a bar - arrogant bastard - it's not rocket science after all! :P)
Although I've often wondered if top class Snooker players have as good eye/hand coordination.
I have rarely seen a field where so much potential good has spent so long in coming to being actually useful. It's "hands" can be made smaller and its motions faster. Best of all making more surgeons is a production engineering problem, not a decades long educational one.
As regard to how "good" a robot surgeon is let's try for the basics.
1) Narrow range of very common procedures. The kind surgeons don't want to be stuck doing.
2)With as good a survival rate (for the given class of patient) as a good surgeon.
3) Ideally operating at least as fast as a good surgeon, but 2 is more important.
4)Able to perform as many procedures as a good surgeon in the same time frame (week, month, year). So no ground-up rebuilds after one.
5)Cheaper to purchase and support than a first rate surgeon on a yearly basis. Around £70K pa for a General Surgeon (the kind you'd prefer not to have operating on you if it's in any way specialised) in the NHS.
In europe high throughput cytology systems will image your cell smear test with sub-wavelength resolution, at multiple wavelengths, with automatic tracking of focus and distortions of the slide, and process the images through a knowledge base of millions of other scans.
In America a doctor looks at <1% of the slide by eye through a microscope bought when first starting out 30years ago, that has never been cleaned, aligned or calibrated. And uses their judgement to decide on the statistical variation of abnormal cells. Automated cell screening systems are banned.
In one of these scenarios the detection rate is indistinguishable from chance.
In America a doctor looks at <1% of the slide by eye through a microscope bought when first starting out 30years ago, that has never been cleaned, aligned or calibrated.
Silly - doctors are way too expensive. Nurses do this.
Presumably, a robotic surgeon would be able to use its cameras and other sensors to ensure that all sponges tools are accounted for prior to closing up the patient.
Also of concern for a Level 5 (unsupervised) robotic surgeon would be this scenario:
a patient is having his appendix removed. Difficulty factor: said appendix has burst. The Level 5 robosurgeon must be able to spot this situation and perform the extra cleaning/disinfecting of all the nooks and crannies of the surrounding abdominal cavity to ensure the patient doesn't die a few days later of peritonitis.
True.
Actually an obvious problem for this is that a very small percentage of people have their organs reversed in the body cavity.
Now should a robot be programmed to spot this or should it just not be allowed to operate on such patients (assuming they can be detected before surgery)?
A full blown autonomous surgical robot will probably need
Real time vision
Task adaptation (what happens when when there are deviations from task description in database EG scalpel blade snaps. Unexpected high level of bleeding, suggesting unexpected or unplanned cuts to patient)
Route planning to get from incision/body orifice to work site.
Along with a high degree of dexterity, preferably with standard surgical instruments.
Defects in craftsmanship of robot, improper maintenance of robot, robot firmware writers with absolutely no sense of secure coding, malware, hackers, BSOD/WSOD/Guru Meditation Error, the robot surgeon that chopped up Farrah Fawcett in "Logan's Run".
Current robotic surgeons are following trajectories designed by human surgeons, they're essentially extensions of those skilled humans. They're useful because they can accurately follow trajectories, they've can be made with a lot more degrees of freedom than a human wrist and they don't have off days.
What people tend to overlook when writing about robots is that they're not mechanical humans, they're machines that have to be told what to do, how to do it and what to do when things go wrong. Once told they'll work tirelessly and accurately, something few humans can do. To the casual observer the "machines are taking over". To the engineer its not only doing a job but providing good eating for engineers and technicians.
Incidentally, you wouldn't believe the amount of testing and certification required to get equipment cleared for medical use.
Its possible that robotic surgeons can take over some routine decision making in the future. Some orthopedic procedures like knee replacement already use computer aids to help determine how the proestheic should be fitted and aligned. Ultimately, though, the human's making the decisions for now.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
