Robotic and computer-aided surgery in the operating room of 2010

Citation

Davies, B.L. (2001), "Robotic and computer-aided surgery in the operating room of 2010", Industrial Robot, Vol. 28 No. 5. https://doi.org/10.1108/ir.2001.04928eaa.002

Robotic and computer-aided surgery in the operating room of 2010

Robotic and computer-aided surgery in the operating room of 2010

Keywords: Robots, Surgery, Computers

The "Mechatronics in Medicine" Group at Imperial College has been researching into robotic and computer-aided surgery (CAS) for the last 15 years and so it is perhaps not surprising that my views of changes in the operating room of 2010 are a reflection of this experience.

Currently, CAS systems are primarily used for navigation purposes to assist the surgeon, by locating the position of tools with reference to target bones or tissue that is to be biopsied, cut or cauterised. The navigators are usually camera-based tracking systems that are used to display tools and tissue locations on a computer screen. The tissue has usually been pre-operatively imaged and modelled in 3D, with sub-millimeter overall accuracy. It should be emphasised that the CAS systems are an aid or a guide for the surgeon, who has to rely on judgement and is clearly in charge of the procedure. The system cannot constrain, but only gives warnings when a wrong action has been taken or when nearing critical areas. The use of CAS systems, particularly for minimal access surgery, will become much more widespread in the near future.

Robotic surgery systems, however, can guide and constrain to safe regions and are generally slightly more accurate than CAS systems. The relative merits and the use of the more costly and complex robotic systems, compared with CAS, will likely become more apparent over the next decade. Orthopaedic applications of robots and CAS systems are currently predominant, because bones can be treated as fixed objects that do not change shape as they are pressed or cut, unlike soft tissue such as is encountered when removing the gall bladder. The main advantage of robots is that they can cut hard bone without bouncing off and damaging adjacent soft tissue. CAS systems, however, require traditional fixtures to guide the cuts (often inaccurately), with the result that the pre-operative plan may not actually be achieved. Thus, in the next decade, it is the accuracy and ability to constrain which will result in robots being used in applications where these benefits are crucial, such as in spine surgery and in small joints such as the ankle, wrist, fingers, etc. These orthopaedic robot systems are likely to be much smaller (and cheaper) than those currently used for hip and knee surgery. Most commercial orthopaedic robots are used as "automated" systems. That is, the patient's limb is securely clamped, preliminary manual exposure of the joint is made, and the robot is then placed at the start position and carries out the pre-determined fixed machining sequence. The surgeon simply observes, to ensure all is well, and presses the stop button if something unexpected occurs. The author believes that this strategy will be superseded by a more hands-on approach to robotic surgery, such as in the Imperial College "Active Constraint Robot" (ACROBOT) currently being used for knee surgery. Here the surgeon holds a force-controlled lever on the end of the robot and moves the rotary cutter over a permitted region under servo-assist. When motion is attempted into a no-go area, (e.g. one which would cut into ligaments), the servo-system transitions from force control into a position control mode, which allows motion along or inside the constraint boundary whilst preventing motion into the restricted region. This active constraint concept provides good synergy between the surgeon, with his experience and judgement, and the robot which provides the constraining and accuracy functions. The hands-on approach means that not only is the surgeon able to adapt readily to changing circumstances, e.g. reducing the cutting rate when meeting hard bone, but the robot can be regarded as just another "intelligent" tool used by the surgeon. This means that there is no doubt that the surgeon is in charge of the procedure, and not the robot supplier, thus reducing possible litigation problems. It is the author's view that such a hands-on approach to robotic surgery, possibly with automated sub-routines for repetitive motions, will be common in the next decade.

Telemanipulator (master-slave), robotic systems have been used for surgery in critical soft-tissue applications, such as cardiac surgery, with minimal access holes for entering the tools rather than the conventional open-heart surgery. This is not quite "hands-on" surgery, since the surgeon operates the master generally in the OR alongside the patient, whilst the slave is observed by a medical assistant. The visual displays are magnified and are of very high quality, often 3D. There is currently very little force or haptic feedback implemented apart from gripping forces, and so all of the feedback relies on vision. The surgeon generally inputs commands at the Master by manipulating sophisticated joysticks. These are mimicked by the slave, often with position scaling, to carry out the actions. Thus the surgeon can manually track soft-tissue as it moves and deforms, without using pre-set trajectories.

Future developments will see the introduction of haptic feedback with a range of scalable touch sensors and the ability to carry out small automated subroutines (such as to automatically stitch along a defined path). The current high costs will also need to be reduced in order to justify the application of both telemanipulators and standard robots to a wide range of less critical, minimally invasive procedures.

One of the major difficulties facing medical robots is the question of safety. Since industrial robots are generally recommended for use away from people, medical robots require additional features to ensure their safe operation adjacent to patients and medical personnel. However, each additional safety aspect will increase the cost and complexity of the system, and so there is an urgent need for international standards to be proposed to ensure adequate safety for the tasks without excessive costs.

The next decade will also see the development of a range of intra-operative imaging systems that will allow the tracking of tissue as it moves and deforms, e.g. to track "brain-shift" when resecting deep-seated tumours in neurosurgery. It will then be possible to track tissue, e.g. using small 3D ultrasound probes placed under the skull flap, and superimpose, (in real-time), these locations onto the pre-operative MRI/CT images. Low Tesla MRI systems can also be used in a slightly modified OR, and then moved to the floor away from the patient, thus allowing the use of normal ferrous tools and conventional metal operating tables. These real-time imaging devices will become commonplace, thus providing intra-operative imaging at crucial phases of the procedure.

The addition of so many complex systems to the OR currently results in a mess of tangled wires, non-standard communication protocols and equipment requiring experts to set up and administer. Given the worldwide difficulties in keeping the same support team in place for any length of time, simple set-up procedures are essential. The future should see the integrated OR, e.g. with strengthened ceilings that permit robots, CAS navigators and imaging systems to hang down over the OR table. Communication protocols will require standardising and the use of expert systems to simplify set-ups, so that the current cohort of highly trained support staff is no longer necessary.

The next decade should see every hospital have robotic and computer-aided surgery systems, permanently located in at least one specialist operating room within the hospital.

Brian L. Davies