The future of Robotic surgery

Is using robots the natural progression of surgery or just a new expensive technique with the same results as human-operated (conventional) laparoscopic surgery?

The first questions that come to mind are, do we really need robots to help us perform our surgeries? Am I willing to hand over my life or the life of a loved one to a bunch of wires and circuitry? Aren't human beings superior to machines in judgement calls and decision-making? Well, yes, maybe in “ideal circumstances”, we are superior. For example, after a good sleep, feeling rested and stable, not emotionally involved, and operating in the first few hours of the morning (when the surgical team have the highest concentration levels).?Then yes, in ideal circumstances, we are superior. But that's only the case about 25% of the time.

In the USA, mistakes by clinicians lead to more than 200,000 US deaths per year, and I imagine that a large percentage of errors were due to doctors and surgeons being tired. An immediate advantage is that robots run on electricity, thus don't get tired, require sleep, or need coffee as humans do. Furthermore, robots with machine learning capabilities can use the combined experience of thousands of surgeons to bring you the best skills and techniques while filtering out all the mistakes coming from fatigue, emotions, panic, and uncertainties that come with being a human surgeon.

As a surgeon, I believe that robotic surgery is the natural, necessary, and absolute progression. ?However, we need to clearly differentiate between robotic-assisted surgery (RAS), (also called supervised autonomous surgery) and robotic performed surgery (RPS). Although RPS is still in animal trials, RAS has been adopted in many hospitals and shows excellent results, but neither is considered the standard of surgery (yet). However, in 10-20 years from now, I'm sure there won't be any other standard.

Why do I believe RAS will be standard in 2031 and RPS in 2041?

Let's first start with some history. Did you know that robots are not as new as many people realize? The first robot patent was filed in 1954 by George Devol for an industrial robotic arm. In 1958, NASA was launched, and it accelerated robotics further. In 1985, the first medical robot, the PUMA 200, began supporting surgeons CT guided brain biopsies. Since the first robots from the 1950s, innovative robotic advancements have occurred, making robots safer, more reliable, versatile, and less expensive than before.

For the robots to become fully autonomous (the switch from RAS to RPS), they must first learn from experience during the supervised autonomous surgery. Some of that experience is already there, with a few million procedures already done with RAS in multiple disciplines, including General Surgery, Gynaecology, Urology, and Orthopaedics. The majority of these have been conducted in the past 20 years, and the recorded experience is still on the hard drives of the assisting robots. I think in the future, it should be conducted in read-only real-time mode; so that computers can analyze the exabytes of data pooling from multiple surgeons performing similar surgeries in different global locations. Therefore, providing machine learning to create the perfect pathways and steps to perform the perfect surgery, according to the pre-op data, intra-op procedure, and post-op outcomes. I appreciate that this would be a massive project with so many potential complications regarding privacy issues, companies' patents, and intercorporate obstacles; but, with the sheer number of procedures being performed yearly, I am sure that the data gathered will accelerate the transformation of RAS to RPS in a way no one can yet imagine. Thus, as I see it, the current intent is not to replace surgeons by 2031, but to improve their results by complementing human skills with enhanced vision, dexterity, and machine intelligence while simultaneously allowing the robots hands-on (in a matter of speaking) learning from the best global surgeons. So, we can create the perfect surgeon, which is perfect every time all the time.

This changes with time as with RAS the surgeon is actually teaching the robot; the data from the operation is sent and stored in the cloud and then fed into an ML enabled program to be later sent back to the robots’ local mind to move closer to autonomy in performing surgeries.

As more surgeons use RAS, this would result in multiple surgeons teaching an ML mind in the cloud and thus combining their experiences together.

The aggregated data, fed into its cloud mind, from thousands of surgeons around the world would learn faster than any human could. However, I have predicted 2041 for RPS as the data that is used to train the ML could lead to results that perpetuate—and even exacerbate current issues in surgery. The reason why is that the machines do not and cannot verify the accuracy of the underlying data they are given. Rather, they assume the data are perfectly accurate, reflect high quality, and are representative of optimal care and outcomes. Hence, the models generated will need to be optimized to approximate the outcomes that we are generating today. It is even harder to address AI-generated disparities because the models are largely "black boxes" devised by the machines and inexplicable, and far harder to audit than our current human health care delivery processes.

The creators of the Smart Tissue Autonomous Robot (STAR) indicated their recent findings and presented them in the 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017). Their experiments demonstrated that their robotic system can be more precise than expert human surgeons performing the same tasks. The study was conducted using three kinds of pig tissue, namely fat, muscle, and skin. Using its advanced 3D imaging systems and its near-infrared (NIR) camera, STAR was able to demonstrate that it could make accurate incisions in all three types of tissue, using tiny infrared markers laid out on the tissue by the researchers for guidance. It is harder than it seems, as these tissues are not consistent in texture and can, therefore, result in the electrosurgical cutting tool slipping and veering off-track.

STAR was able to perform the task on all three tissues well, thanks to its instruments, which are equipped with sensors that can precisely gauge the force needed to consistently cut into different types of surfaces, constantly adjusting for variables like movement and texture.

STAR was then pitted against human surgeons to undertake the same cutting task, this time on pigskin. The first exercise involved cutting a straight line measuring five centimetres. Some of the surgeons used different surgical methods to achieve the same result (i.e., open, or laparoscopic surgery), but overall, the robot's incisions were closer to the desired length, deviated less from the ideal cut line and had less damaged flesh around the cut.

The next segment of the experiment involved STAR removing an imitation tumour, created with clay, out of a section of pig fat. The area was once again delineated with markers that are visible to the robot's near-infrared camera, but to make the task more difficult, a thin piece of skin was layered on top in an attempt to confound the robot's visual tracking system. Nevertheless, the robot was able to make the cuts with impressive accuracy, well within the four-millimetre margin around the tumour, as set by the researchers.

Even if robotic surgeons don't completely replace their human counterparts, robot-assisted surgeries would nevertheless be good for human doctors too, as they help to alleviate some of the physical strains that hours of performing these delicate procedures might bring upon a human surgeon. These include eye strain, hand, neck, back and leg problems, and carpal tunnel syndrome, not to mention the mental fatigue — all of which can cause a surgeon to make mistakes or take on retirement much sooner.

These problems of physical strain are even more pronounced when it comes to doing minimally invasive procedures such as those found in laparoscopic surgery, where tiny incisions must be made, and surgical procedures are done with the help of small video cameras and instruments that are inserted into the patient's body. Since they are often looking at a video screen for these procedures, human surgeons do not have that direct "hands-on" experience that would be helpful during the operation. For a robot, this would be less of an issue, meaning that it's quite likely we'll be seeing more autonomous machines in our operating rooms in the future.

Some tasks, such as suturing and valve repair, tend to be the tasks that surgeons consider boring and repetitive, so these are ideal as the first surgical tasks to be outsourced to a partial RPS.

Considering all what we mentioned of benefits, the success of animal trials in RPS, the reduction of human errors, adoption of robotic surgery becomes an almost absolute necessity for any hospital or medical institute that wants to survive the next few decades. So, what's the hold up? Why aren’t all hospital using robots? ?The answer is simple, money. The benefits are clear; however, any piece of equipment needs to be costs effective to either save money or make money for the hospital.

Keeping that in mind, starting up a robotic surgery service in a hospital is a complicated endeavour, the simplest part of which is buying the 2-million-dollar machine. I've known many hospitals who purchased the machine and left it collecting dust for years, either for technical problems, running cost, insurance denials, lack of surgical expertise, or dysfunction and miscommunication between different hospital departments.

The most important requirement for starting a robotic surgery service is the full involvement and investment of the hospital from the CEO to the insurance coordinators. Surgeons need to be involved to know what kind of robot they need and what kind of surgeries they are going to perform. RCM needs to be involved in budgeting and to know that you're going to get enough patient flow. Theatre management needs to be involved to set-up dedicated theatres for the massive machine. Biomedical and IT should be prepared to deal with any problems that may arise, and they will need special courses. Marketing should be involved before the service is even started. The company selling you the robot will help assess your hospital and will help you set up the service, but you still need a full committee from your hospital to make sure that everything goes according to a strict timeline, and you need to set up goals that are realistic, precise, and achievable for this venture.

In summary, using robots, both RAS then RPS, is the natural progression of surgery; however, without excellent management both will be a very expensive investment that could take money away from alternative patient care.