Medic-aid

The operating theatre, for years the sole preserve of highly skilled surgeons with steady nerves and steadier hands, is fast becoming home to a new generation of medical assistants with even steadier hands, and no nerves at all: medical robots.

Advances in surgical knowledge and techniques, coupled with improvements in robotics and imaging technology, are leading to the increasing development and use of robotic assistants. From robot medics on wheels to specialised systems designed to complement and enhance a surgeon’s skill, robots offer tremor-free, minimally invasive procedures that could potentially result in less risk of damage, a speedier recovery time and a shorter hospital stay.

But some in the medical world have voiced concerns that technology will relegate surgeons to the role of a minor technician. Do surgeons really want to spend years training — only to find the closest they can get to a patient is behind a joystick?

Prof Brian Davies of Imperial College, London is the world’s first professor of medical robotics and a vocal champion of this kind of system. He believes that the perception that robots will de-skill the surgeon is not only unfair, but is holding back development and uptake of the technology across the world.

‘Surgeons are trained to be hands-on with the patient and be entirely responsible for the procedure,’ he said. ‘There has been a sense [when using robotic assistance] that “I’m here, sitting with my finger on the emergency-off button”, and that is the only involvement the surgeon has.’

Davies believes instead that a system that complements the surgeon represents an opportunity for less skilled surgeons to carry out the most demanding surgical procedures — surely a bonus for patients as waiting lists extend.

Davies’ belief that robotic accuracy and hands-on surgical skill need not be mutually exclusive has resulted in Acrobot. Developed by his spin-out firm, the London-based Acrobot Company, this knee-surgery robot was designed to allow the surgeon to move a cutter around a patient’s knee to remove bone.

Using a concept called active constraint, the robot system can assist or resist the surgeon if it senses it is cutting into sensitive areas of the knee, thereby minimising possible damage to surrounding tissue or ligaments.

Essentially, once the constraints have been computed, the surgeon selects where to make the best cut and the robot automatically goes into position. The surgeon can then move a rotating milling cutter at the end of the robot to remove sections of bone.

As this is being done, an image appears on the screen in a number of views, together with a depth indicator that shows how deeply the bone has been cut. These cues, and the variable stiffness and constraints, allow bone to be cut accurately with no damage to surrounding soft tissue.

Earlier this year Acrobot took part in a clinical trial at London’s Middlesex Hospital, and has since been approved by the medical devices directorate of the EC. This trial took the form of a minimally invasive knee procedure, with 15 patients operated on by Acrobot and 15 with conventional surgical techniques. The result revealed that even though both had used the same planing procedure, the robotic assistant gave a far greater accuracy — within two degrees of alignment, while the surgeon was as much as nine degrees out.

Davies told The Engineer that he expects the system to be used by a small number of UK hospitals in the third quarter of 2006, and be rolled out across the country by the middle of 2007. He’s also keen to enter the highly lucrative American market. ‘We will be going for approval by the Food and Drug Administration because the US is the biggest market of them all. But we need to ensure good data in Europe to minimise FDA time and requirements. That would not happen until 2007,’ he said.

Beyond the operating theatre, Davies has also been investigating the use of Acrobot for training purposes. Its semi-active nature means that it can be linked with force feedback systems and used to help trainee surgeons experience a more realistic sense of touch when training.

For example, in the case of a knee arthroscopy Davies said current methods — involving a plastic bone — are a poor reflection of real life. The limitations of existing training techniques can be bought into stark and grisly focus the first time a surgeon has to carry out the procedure on a real knee.

Davies’ company has also been working with Corin, in Cirencester, which supplies orthopaedic prostheses for hip replacements. While conventional hip replacement entails removing the whole head of the femur, the idea is to reshape the head of the femur and sit a prosthetic ball on top. Acrobot has signed an agreement with Corin to develop a device to do this using its own navigation system and a roll-out of this technology can be expected early in 2006, said Davies.

But the biggest trend in the world of medical robots, he explained, is the development of smaller, cheaper systems that are designed for specific operations. He pointed to the work of Mazor, an Israeli company that has created a small robot for drilling holes into the spine.

The Spine Assist system, which recently received FDA approval, has undergone a number of clinical trials. Perhaps most notably, the device was used last month to precisely locate a tumour in a patient’s spine. The surgeons involved in the trial concluded that the device enabled a complete excision of the benign thoracic tumour with minimal removal of healthy bone.

The majority of medical robots being used or poised to enter the surgical arena are of the master-slave or tele-operated variety — where a skilled surgeon will operate a robotic system. Perhaps the most famous of these is the Da Vinci surgical system developed by US company Intuitive Surgical. One of the first of its kind to enter service, this enables surgeons to get closer to the surgical site than human vision will allow, and allows them to work at a smaller scale than conventional surgery permits.

But Davies, while impressed by the core technology in the Da Vinci system, is less convinced of its cost-effectiveness: ‘This is $1m (£500,000) worth of kit, so it is not justified for gall bladder removal. Cost-effectiveness is a big problem for this area. When you look at all aspects of gall bladder removal that are not critical, it is difficult to make a cost justification for such a system,’ he said.

These sentiments were backed up Dr Patrick Finlay, chief scientist and director of robotics manufacturer Armstrong Healthcare. ‘It is a very impressive piece of engineering,’ said Finlay. ‘What makes it so good is the highly dexterous wrist inside the robot, so you can manipulate it to allow suturing, stitching inside the patient with good dexterity.

‘But the downside is it is very expensive and bulky and, in terms of cost-effectiveness, difficult to justify. In terms of outcomes, there is no evidence that it can do anything better than a surgeon can.’

Finlay’s company has developed an image-guided neurosurgery robot designed to provide greater accuracy while reducing the duration and complexity of procedures. Known as PathFinder, the system has capitalised on the huge advances made in medical imaging over the previous decade. It uses CT and MRI scans to operate on intercranial lesions without having to build a stereotactic frame — a form of scaffold that clamps to the head to appear on the CT scan to triangulate a position.

In essence, the robot uses the CT scan like a map to position the probe to an accuracy of 1mm. Once in place it can then insert the instruments or allow the surgeon to take over. PathFinder is at the beta phase of development, with clinical trials being held at neurological centres in Nottingham and Dundee, and Finlay hopes it will go into commercial production in 2006.

Another device from Finlay’s lab that has already obtained FDA approval and is in use is an ‘intuitive’ camera control for endoscopic surgery. Known as EndoAssist, the system consists of an intelligent robot arm with a camera fitted to it. Instead of a team member holding the endoscope, its movements are controlled by the surgeon’s head. If the surgeon tilts his/her head to look left, EndoAssist will register this and move it accordingly. The benefit of this, said Finlay, is that it gives control back to the surgeon and allows for more delicate procedures.

He added that, with minimally invasive surgery on the increase, there is a particular demand for this kind of technology. ‘As more operations become accessible using minimal invasive techniques, the need for robots becomes greater, because the operations last longer and so the problem of tremor is more apparent,’ he said.

As in other areas of robotics, the technological reality of medical robots is generally some way behind the sci-fi vision of fully autonomous, walking, talking android doctors. But one man who is perhaps doing more than any other to push forward the technological boundaries is Imperial College’s Prof Ara Darzi, the robotic surgery pioneer behind the UK deployment of the Da Vinci robot. Darzi was the latest recipient of the International Hamdan Award for medical research excellence for his work in minimally invasive surgical technology using robotic surgery.

One of the most high-profile examples of Darzi’s recent work is the Remote Presence 6 (RP6), nicknamed Sister Mary after St Mary’s Hospital, London where it was trialled. A ‘remote medic’, this robot is designed to carry out ward rounds for consultants. Patients can communicate with the doctor, stationed in his office or elsewhere, through a video screen fitted on to the unit. According to Darzi, the system is particularly useful for medics to ‘visit’ patients who are being kept in isolation to prevent the spread of diseases such as ‘superbug’ MRSA, which is resistant to most antibiotics.

Darzi said he is working on a number of advanced procedures to improve robots’ interface with operators and other systems. One potential application is a smart gown that could be used to measure a patient’s blood pressure, pulse and temperature. Darzi said that the idea is that the gown will take these and automatically send them to a robot such as the RP6.

But while Darzi is full of enthusiasm for the benefits this kind of technology can bring, this is tempered by what he sees as a somewhat moribund commercial climate. ‘The biggest threat to the robotic future is the lack of competitiveness in the market,’ he said.

The take-over merger of Intuitive, the company responsible for the Da Vinci, and Computer Motion has, he felt, had a particularly negative impact on the future of robotics, creating a less creative and innovative climate in the industry.

While Darzi advocated that further funding is essential to the future of robotic surgery, he is not confident that it will be immediately forthcoming.

Davies agreed: ‘The UK has always been quite slow compared to the rest of Europe. In some ways the NHS as a purchaser is very risk averse so it is not at all welcoming of innovation,’ he said.

Clearly there is some way to go before science-fiction’s dream of an all-knowing, super-fast ‘robodoc’ is realised, but there are plenty of areas of development to get excited about.

For instance, Darzi, Davies and Finlay all anticipate that improvements in imaging technology will become increasingly significant. They claim that intra-operative imaging, which enables robots to modify targets as they work, is certain to prove a powerful tool in future surgery. ‘Imposing pre-operative images and augmenting them on top of video images during surgery, so you can see beyond the surface, allows you to see important nerves that are not visible,’ said Darzi. ‘It adds another dimension. It is very exciting and novel.’

Finlay added that further into the future, advances in imaging systems are likely to lead to even more astonishing developments with, for instance, surgical robots able to operate on moving organs such as lungs or a beating heart.