In the early 1980s, general surgeons started to experiment with a new type of surgery—minimally invasive surgery (MIS) in an attempt to minimize the trauma of surgery for their patients. This certainly worked well in certain circumstances, and the extremely good results with MIS cholecystectomies bear witness to the success of such surgery. However it soon became apparent that limiting the surgeon's access to what he or she could see and reach through a few ports brought with it a whole new set of problems. First, introduction of the endoscope forces the surgeon to work while looking at a video image rather than at his hands, which disrupts his usual hand–eye coordination. Second, conventional endoscopes use two-dimensional vision, resulting in loss of the depth perception usually afforded by our normal binocular vision. Third, the instruments for MIS are introduced into the body through ports. These ports act as pivots, causing the instrument to move in an opposite direction to the surgeon's hands. Finally, the port in the body wall constrains instrument motion in two directions, so that the functional tip of the instrument has fewer degrees of freedom—from the usual six degrees of freedom to four. In an attempt to overcome these problems, people decided to look at different ways of controlling the endoscopic instruments. Autonomous robots had been around for some time. These can be programmed to move with great precision and speed, but they are unable to perform activities requiring complex physical interaction or to respond to unplanned events. Master-slave manipulators, also known as telerobots, place a human in the control loop, thus taking advantage of the human's cognitive and sensorimotor skills and the robot's fine motor ability. Telerobots have been used for years to handle hazardous waste and defuse bombs. A new type of telerobot was obviously necessary to work in the surgical environment with a whole new set of standards in terms of precision and scale of movement. In addition, telerobots had been designed to protect themselves, but now the patient would need protection from injury. In the late 1980s, researchers at Stanford Research Institute (SRI) developed a prototype system with funding from the National Institutes of Health. This SRI system combined advances in remote manipulation with basic force feedback, stereoscopic imaging, multimodal sensory feedback, and ergonomic design. Funding was also provided by the Defense Advanced Research Projects Administration (DARPA), which sought a system that would enable surgeons at a remote hospital to operate on troops injured on the battlefield. Other sites involved in this type of research at this time included Massachusetts Institute of Technology (MIT), IBM's Watson Laboratory, NASA's Jet Propulsion Laboratory (JPL) working on an opthalmic robot, and Computer Motion's automatic endosopic system for optimal positioning (AESOP). Intuitive Surgical was founded in 1995 by Dr. Fred Moll, a surgeon with experience in founding two previous surgical companies, Rob Younge on the technical side, and John Freund on the financial side. Their aim was to produce a medical telerobot that would be at the forefront of computer-aided surgery and create a surgeon–robot interface that was transparent to the surgeon so he or she could use surgical skills in a natural and instinctive way. To this end, they hired a group of scientists and engineers with diagnostic and robotic experience. They licensed technology and hired engineers from SRI, MIT, and IBM. By the spring of 1996, after only 4 months of work, the first prototype was built for animal trials (Fig 1). The prototype had a wrist enabling six degrees of freedom inside the body and stereoscopic vision. One year later, in the spring of 1997, a second prototype design was tested on humans in Belgium in general surgical procedures. The spring of 1998 saw the alpha prototypes of the daVinci (Fig 2) in use in Paris and Leipzig, Germany for cardiac procedures. An FDA trial for laparoscopic indications was conducted in Mexico City in the summer of 1998. Four surgical teams performed approximately 100 laproscopic daVinci cases and 100 control cases for two procedures, laparoscopic cholecystectomy and Nissen fundoplication. The CE mark was received in 1999, and during that year 12 systems were sold. The year 2000 saw 28 systems sold and FDA approval for laproscopic use. FDA approval for thoracoscopic use came in March 2001 after United States clinical studies at the Ohio State Medical Center. The challenge for Intuitive's engineers was to develop a system that creates “an immersive operating environment for the surgeon by providing both high-quality stereo visualization and a man-machine interface that directly connects the surgeon's hands to the motion of the surgical tool tips inside the patient's body.”1Falk V McLoughlin J Guthart G et al.Dexterity enhancement in endoscopic surgery by a computer controlled mechanical wrist.Min Inv Therapy Allied Technol. 1999; 8: 235-242Crossref Scopus (56) Google Scholar This challenge is met by the DaVinci system, which comprises two major subsystems: the surgeon's console and the patient side cart. The surgeon's console houses the display system, the surgeon's handles, the surgeon's user interface and the electronic controller (Fig 3). The surgeon sits at the control console looking at an image that is displayed as if it were situated over his or her hands. Each movement of the surgeon's handles, or master is translated in real time to movements of the instrument tip, or slave. These movements can be scaled from 1:1 to 1:3 while the control system is also able to filter out surgeon tremor, making the instrument tips steadier than the unassisted hand. The combination of the motion scaling and filtering and image magnification makes delicate motions easier to perform than in conventional endoscopic techniques. The controller is also able to translate the surgeon's movements exactly resulting in a movement by the surgeon to the right being replicated by a movement to the right by the slave. All the time that this is happening, the movements of the slave and master are being checked and correlated in the x, y, and z axes at more than 1300 times per second. Incorporating more than 250 megaflops of processing power makes it possible to bring the visual and robotic frames of reference into precise registration—a key element in giving the surgeon a sense of immersion in the work. The system controller also permits clutching (indexing) between master and slave, smooth control at workspace limits, and gravity compensation. The master input device has a large workspace that accommodates the full range of motion required by the surgeon seated at the control console. Its low mass and friction and precise tracking of motion enable exacting commands to be transmitted to the robot. Force feedback lets the surgeon feel large contact interactions and indicates robot workspace limits. A vital initial component was the development of a superior three-dimensional (3D) visualization system. Early experiments by the development team showed that high-quality visualization was the essential first step in providing a system that could delicately manipulate tissue, particularly where force reflection for delicate interactions is not available. In addition, the counterintuitive nature of conventional MIS, with the instrument movements in the opposite direction from the hand movements, had to be corrected. Images are captured from within the patient by a specially designed endoscope with two totally separate optical trains. The camera mounted atop the endoscope has two independent three-chip charge-coupled device cameras that deliver pictures with 800 lines of resolution and a signal-to-noise ratio exceeding 62 dB. These images are displayed on two medical-grade cathode ray tube monitors, each displaying slightly different image to each eye giving 3D vision with a relatively wide stereo separation. This contrasts with other 3D systems that use a single optical train and off-axis imaging to provide 3D vision with relatively narrow stereo separation. This high-resolution 3D video imaging and display system provides the surgeon with an exquisitely clear and bright view of his or her work. The optical system minimizes geometric distortion across the field of view, enabling stereo image fusion even near the edges of the image. To minimize chromatic distortion, the system also provides a highly accurate color rendition. The system projects the image of the surgical site over the surgeon's hands, via mirrored overlay optics, restoring hand–eye coordination and providing a natural correspondence in motions. During a procedure, the position of the camera mounted on a robotic arm can be adjusted to provide the best view of the surgical site. The camera also provides magnification, enhancing the surgeon's ability to perform fine tasks and improving visual feedback. The surgeon's tool handles are serial link manipulators designated as the masters. These masters act both as high-resolution input devices, reading the position, orientation, and grip commands from the surgeon, and as haptic displays, transmitting forces and torque to the surgeon in response to various measured and synthetic force cues. The user interface at the surgeon's console consists of foot switches and buttons that allow the surgeon to control the system throughout the surgical procedure, as well as a variety of other mode selection and initialization switches. This interface allows the surgeon to control the endoscope from the surgeon's console, to reposition the masters in their workspace, to focus the endoscope, and so on. The last major component of the surgeon's console is the electronic controller. Speed, reliability, and fail-safe system operation drove the design of the electronic controller. This custom-designed control computer is capable of fully interconnected control of 48 degrees of freedom at update rates exceeding 1000 cycles per second. It can read up to 48 encoders and 96 analog input channels in real time while driving output through up to 48 digital-to-analog conversions. The heart of the controller is a parallel floating point digital signal processing architecture with a peak computational power of 384 Mflops and a sustained processing power of 128 to 256 Mflops. Surrounding the computer engine of the controller is a network of 24 microcontrollers and integer DSPs performing data transfer and health watch-dog functions. Redundant sensors, hardware watch-dogs, and real-time detection ensure fail-safe operation of the controller in all of its states. The patient side cart consists of a fixed base with three passive multilink arms mounted to it (Fig 4). Each arm holds a slave manipulator, two manipulators drive the tools, and one controls the camera. To overcome the limitations of the instruments used in conventional minimally invasive surgery, the engineers at Intuitive had to come up with a new set of tools to deploy an instrument into the patient's body that was capable of allowing the surgeon seven degrees of freedom (three for translation, three for orientation, and one for grip)—what he or she is used to in conventional open surgery. From a clinical standpoint, a tiny mechanical wrist called the EndoWrist (Fig 5) is a key component of the Intuitive system. The EndoWrist is the component that gives the surgeon the ability to reach around, beyond, and behind delicate body structures, and is connected to the rest of the system by sophisticated mechanical cable transmissions. Its motion is monitored by the computer, so that the control algorithms can translate the surgeon's motions to the robot's wrist. The Endo-Wrist provides four degrees of freedom of movement inside the patient (three for orientation—roll, pitch, yaw—and one for grip). The large range of motions or translations—up/down, left/right and rotation—are provided by the arms of the patient side cart. The computer translates the surgeon's open-surgery hand movements into scaled-down movements of the instruments, with the same orientation of each movement. The EndoWrist delivers any angle of movement directly to the surgical site. The system is capable of applying a fraction of an ounce of force for delicate suturing up to several pounds of force to retract large structures. The tools are the most distal components of the patient side system. They each have the EndoWrist with four degrees of freedom, and are fully sterilizable instruments, including those required for needle grasping, cutting, cautery, and clip application. All of these attach interchangeably to the two tool manipulators. The tools have been developed to look like regular instruments, to help the surgeon feel comfortable with standard instruments. Sterilization is accomplished using standard procedures that have been reviewed and approved by the FDA and European regulatory bodies. Unlike most master-slave robots, the DaVinci requires an assistant to work alongside the slave or patient side cart. This requires very safe and human friendly engineering on interfacing with the slaves, where the nurses are required to change instruments on the tool manipulators. This usually takes from 15 to 30 seconds per instrument change. Telerobotic surgery also requires fail-safe operation. Most telerobots have simple safety systems that protect themselves in the event of failure. That is because in less complex applications, the robot is the high value item. The Intuitive system must protect the patient first, and the robot second. During a procedure, the Intuitive system monitors itself continuously, and shuts down or alerts the surgeon if a problem arises. Safe and fail-safe operations are ensured by redundant sensors and several levels of system health checking. With robot arms and camera mounted on mobile setup joints, the system can be conveniently brought to and removed from the surgical site. The technology is the result of work by a team of more than 100 engineering, medical, and management personnel, guided by feedback from the surgical community. It embodies a combination of capabilities aimed at enabling radically new ways of performing surgery. Although much of Intuitive's engineering effort has gone into accurately translating the surgeon's movements to the robotic arms, remote duplication of a surgeon's existing skill set is only part of telesurgery's potential. The fact that a computer, rather than hardware, provides the interface between the surgeon's hands and the robotic arms provides another aspect of the system's potential: extending the surgeon's capabilities beyond conventional surgical techniques to add new capabilities that are simply not possible in conventional surgery. Although these possibilities are just beginning to be explored, many of them will undoubtedly lie in miniaturization and microsurgery. The telerobotic system may also overcome one of the most basic human limitations—the availability of only two hands. During telesurgery with the Intuitive system, the surgical team can use “dynamic assignment”; that is, one surgeon can theoretically manipulate one arm, leave it in place, and then switch his or her attention to another arm. In fact, multiple robotic arms that in theory can be controlled by a single surgeon may be used in the future. Alternatively, in team surgery, the use of a surgical robotic arm can be taken over by another member of the team, or several surgeons may be able to operate simultaneously by cooperatively sharing and trading control of the surgical tools. The authors thank Gary Guthart and Ken Salisbury for their help with this manuscript.

Full Text

Published Version
Open DOI Link

Get access to 115M+ research papers

Discover from 40M+ Open access, 2M+ Pre-prints, 9.5M Topics and 32K+ Journals.

Sign Up Now! It's FREE

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call