Robots in the Operating Theatre
In a typical robot-assisted operation, the surgeon is not standing over the patient at all. Instead, he or she sits at a console a few metres away, looking into a viewer that shows a magnified, three-dimensional image of the inside of the patient's body. Beside the operating table, a machine with three or four slender arms holds a camera and a set of instruments, each inserted through a cut no wider than a finger. Every movement the surgeon makes with the hand controls is reproduced by the arms, while a team of nurses and an assistant surgeon remain at the bedside to change instruments and to take over if anything goes wrong. Despite its name, the robot does nothing by itself: it is, in effect, an extremely sophisticated extension of the surgeon's hands.
The idea is older than many people realise. The first recorded use of a robot in an operation took place in the mid-1980s, when an industrial arm, originally designed for factory work, was adapted to hold a needle steady during a brain biopsy. Over the following decade, research groups experimented with machines for hip replacements and other procedures, and military planners became interested in the possibility of operating on wounded soldiers from a safe distance. It was not until around the turn of the century, however, that complete systems became commercially available, and even then they were found mainly in large teaching hospitals.
Supporters point to a range of technical advantages. Software filters out the natural tremor of the human hand, and a feature known as motion scaling allows the surgeon to make large, comfortable movements that the robot converts into much smaller ones: a hand movement of three centimetres, for instance, might move the instrument by just one. The instruments themselves have joints at the tip that can rotate further than a human wrist, enabling surgeons to stitch tissue in confined spaces deep inside the body. Because the incisions are small, patients generally lose less blood, experience less pain and leave hospital sooner than those who undergo traditional open surgery.
Yet the most rigorous studies have produced a more cautious picture. When robotic procedures are compared not with open surgery but with conventional keyhole surgery, in which surgeons operate through similar small incisions using hand-held instruments, the differences in outcomes are often slight. For several common operations, large reviews have found no meaningful difference in complication rates or recovery times. Professor Tomas Arvela, a health economist, argues that this matters because the costs are considerable. 'The purchase price is only the beginning,' he says. 'There are annual maintenance contracts, and many of the instruments can only be used for a limited number of operations before they must be replaced.' He also suspects that some hospitals buy robots partly as a marketing tool, since advertising the latest technology can attract both patients and talented surgeons.
Learning to operate with a robot also takes time. Even experienced surgeons typically need to perform dozens of procedures before their operating times and results match those they achieved with conventional methods, and many hospitals now require trainees to practise on simulators, which record every movement and score the user's accuracy, before they are allowed near a patient. One recurring challenge is the absence of touch, since most systems give the surgeon no sense of how firmly an instrument is pressing on tissue. Dr Lena Marshfield, who trains surgeons in robotic techniques, says that people adapt surprisingly quickly. 'You learn to read the tissue with your eyes,' she explains. 'When you see it stretch or turn pale, you know you are pulling too hard.'
One of the more ambitious promises of robotic surgery is that an expert could operate on a patient in another city, or even on another continent. Several demonstrations have shown that this is technically possible, but routine use remains distant. The main obstacle is delay: a signal travelling over long distances takes time, and if the gap between the surgeon's movement and the robot's response exceeds a fraction of a second, precise work becomes dangerous. Faster communication networks may reduce this problem, but they cannot remove the need for a fully trained team at the patient's side.
The most controversial question concerns autonomy. Researchers have already built experimental systems that can carry out limited tasks, such as stitching a wound in animal tissue, without moment-to-moment human control, and in some trials their stitches were more evenly spaced than those of human surgeons. Dr Imogen Rask, an engineer who designs surgical robots, believes that automation will arrive gradually. 'Machines will be trusted first with simple, repetitive steps, such as closing a wound, long before they are allowed to make decisions,' she predicts. Who would be responsible if an autonomous system made a fatal error, however, is a question that neither engineers nor lawyers have yet answered.
