Dutch researchers have developed a robot for performing complicated eye surgery, with the intent of enabling eye surgeons to operate with increased ease and greater precision on the retina and the vitreous humor of the eye.

The system, developed by Thijs Meenink and Ron Hendrix at the Eindhoven University of Technology (TU/e), The Netherlands, is expected to extend the careers of human eye surgeons by filtering out the slight hand tremors that prevent older ophthalmologists from carrying out delicate operations on the eye. The system could also bring haptic feedback technology to eye surgery; currently, ophthalmologists work by sight alone. 

The robot consists of so-called master and slave units, and is equipped with two joysticks for control. Two robot arms on the slave are fitted with tiny (0.5 mm) forceps, surgical scissors, and drains. The surgeon can switch instruments in seconds—a particularly useful feature given that a single eye operation can require as many as 40 instrument changes.

Movements on the joystick are scaled down so that each centimeter of movement is translated into just one millimeter of movement within the eye. According to the researchers, mechanical tests have demonstrated very low actuation torques, and the ability to make very precise movements.

The system is also designed so that the needles used always enter the eye at exactly the same point, reducing the potential for damage to the eye’s delicate structures.

The device’s haptic feedback feature measures the tiny forces at play during operations, amplifies them, and sends the resulting signal back to the surgeon-operated joystick.

“At the slave robot, we used a commercially available … force/torque sensor, which we placed directly near the instrument,” Meenink told Robotics Trends. “The master is equipped with direct drive motors. And both master and slave are optimized for high stiffness, low friction … to reflect the applied force as well as possible. There is no feeling of resistance, when the haptic interface [of the master] is moved around offline. This clean design will give a solid base for accurate haptic feedback.” 

Eye-surgery robots are not yet commercially available. However, Meenink expects the first functional tests to begin in the near future and for the new robot to come to market in three to five years.

 “Until now, the focus has primarily been on the research part of the project. Currently, we are working on a business plan, targeting to launch a spin-off company in the near future. We are setting up a team of experts in mechanical design, control, and haptics,” says Meenink. 

 “Our first goal will be to perform further tests and to prepare for pre-clinical use. As most or all of the ophthalmic surgeons we meet are very enthusiastic, we see ample opportunities to bring the robot to the market. Nevertheless, many steps still have to be taken, such as [pre-] clinical tests, regulatory approval, and redesign of specific parts according to the results of these tests.” 

Meenink can count on support from TU/e during the testing and certification processes. The university will fund those processes, provide a site for testing, and help with the creation of a start-up company to market the device.

Nevertheless, getting U.S. Food and Drug Administration (FDA) and European approval will be a “time-consuming and costly process,” says Meenink.

“First, we are going to verify this prototype. Based on experience gained by those tests, there will be a second and improved version designed, which will be designed to meet EU and/or FDA requirements.” 

“As we have all the technology to start experiments, it should not cost much to keep going. If we want to build a second prototype, then we absolutely need funding: to realize that one and to pay for the clinical trails. We are still summarizing all costs to get to the second prototype,” explains Meenink. 

Competing designs

There is something of a race on amongst researchers to bring robotic technologies to eye surgery. Earlier this year, for example, a team at the Institute of Robotics and Intelligent Systems (IRIS) in the Swiss Federal Institute of Technology, Zurich, revealed a robotic system capable of delivering tiny payloads to parts of the eye.

In fact, the payloads are so small they can be injected into the eye without anesthetic. The untethered robotic system—which is essentially a tiny magnet controlled by an electromagnetic field outside the eye—could enable new types of non-invasive surgeries and highly targeted drug delivery. For example, drugs designed to treat macular degeneration and retinal vein occlusions could be delivered to specific locations on the retina.

The IRIS-built device has so far been tested on synthetic eyes and the eyes of dead animals, but the team hopes to test it on living animals and then begin human trials.

A team at the Institute of Robotics and Intelligent Systems revealed a robotic system capable of delivering tiny payloads to parts of the eye.]

Meanwhile, researchers at the Computer Integrated Interventional Systems Laboratory at John Hopkins University in Baltimore are working on Eye Robot 2 (ER2) and a robotic assistant designed for retinal microsurgery. Cooperative control allows a surgeon’s hand movements to exactly dictate the robot’s movements. The robot is also designed to eliminate physiological tremor in the surgeon’s hand during surgery.

The team is currently working on a series of membrane-peeling experiments on real eyes with the aim of developing warning signals for the surgeon if acceptable levels of force are exceeded during surgery, and designing virtual models that can be used as the basis for a surgical training set. 

Satisfying regulators

While the technology may be ready, the real race involves receiving FDA and European Union approval for these new devices. 

On average, a high-performing medical device manufacturer spends around 4 to 8 percent of its revenue on new product development, says James Pink, principal legal consultant on medical devices to NSF International, based in Ann Arbor, Mich., and a U.K.-based expert on FDA and EU compliance requirements.

“For many medical device manufacturers, you could say that the majority of the new product development is spent upon ensuring that the technologies are suitable for the design intent,” says Pink.

In the EU, each member state has its own medical device compliance body, which has the power to enable device manufacturers to sell their products in their country. EU countries do share a common commitment to the European Medical Device Directive, however, which outlines the specific steps a manufacturer must take to prove the safety and effectiveness of its medical devices.

It’s a path that requires independent evaluation of the manufacturer’s design, manufacture, and post-market processes to ensure that all elements of the directive (as well as the hundreds of harmonized standards, industry-based standards, and fundamental science and technology issues relating to the product) have been addressed.

“Without knowing the system I would compare the motion, movement, software, and reliability with that of similar machines used in medicine such as linear accelerators and surgical positioning devices, which use similar technologies and require similar precision. The device may certainly not be commercially approved for use by then [2015], but the regulations throughout the world will not restrict a human surgery provided that work has been undertaken to understand the safety risks, and the device has achieved a level of safety that is generally accepted as the state of the art,” explains Pink.

“Many of the pitfalls occur through a lack of understanding the risk of the clinical application, the claim the manufacturer is making, and then not having sufficient validations to substantiate all aspects of hazardous situations you must avoid,” says Pink. He adds, “With a technology like this, it is highly anticipated that repeatability, precision of movement, and achievement of function will be based upon the design of the system and its software. Software validation alone is a complex sequence of understanding the risk, the software architecture, and the risks that a mechatronic/human interface can bring with it.”  

Despite the complex nature of the process, Pink believes that 2015 is a realistic target for the first surgery to be performed on a human eye using the Eindhoven-designed robot.