Funding partially autonomous surgical robotics development

The research association ForNeR's two million euro grant from the Bavarian Research Foundation's Board of Trustees integrates assistive robots for surgery through intuitive interfaces in the operating room. This not only helps patients but also supports operating room staff in performing demanding procedures.

In an interview with MEDICA-tradefair.com, Prof. Franziska Mathis-Ullrich describes the possibilities of the new robot assistance systems.

Prof. Mathis-Ullrich, how could future robot assistance systems support work in the OR?

Prof. Franziska Mathis-Ullrich: Our goal is for robots and humans to work together as partners in the OR. To this end, we are developing the next generation of learning robotic systems and want to introduce assistive functions that actively support surgeons.

The technology is now so advanced that hand movements can be scaled to reduce the tremor of the surgeons. In addition, we can 'look inside' the tissue with new sensors. In our own research, we use continuum robots that do not work with straight and rigid instruments, but with tools that can move around an object like flexible bodies.

What are the special advantages for robot-assisted eye surgery?

Mathis-Ullrich: In micro-robotics, the scaling of movements plays a special role. About 20 percent of the transmitted forces are so small that they are below the human perceptual limit, so surgeons need a lot of experience and skill. Surgical assistance by robots is well on its way to pushing the boundaries of ophthalmic surgery.

For example, proximity sensors that we have previously explored for other surgical applications could be miniaturized. Used in the eye, these sensors could show how close an instrument is to the delicate retina. Force sensors can be used to provide haptic feedback to operators when they are operating on particularly sensitive tissue.

Can partial steps of an operation also be automated?

Mathis-Ullrich: Our goal is to be able to offer various assistance systems in automated form. The eye is an organ that is not very deformable and the dimensions are comparatively similar in all people. It therefore offers a relatively standardized environment that facilitates automation.

To support surgeons, we want to display and visualize the planned positioning of the robot. In the case of autonomously performed surgical steps, the robot must also explain why certain decisions are being made. This is the only way to build trust between the robot and the surgeons, who can retain control and trigger automated procedures in a controlled manner.

These processes should be intuitive and efficient. So is it also about the operating concept?

Mathis-Ullrich: To be able to establish a system with autonomous steps in the clinic, the partnership between system and human must be optimized. The use of robotic systems requires a comparatively large amount of preparation time. The system has to be placed on the patient and the robotic arms with the tools have to be prepared individually.

We want to provide intuitive ways to use systems more efficiently, quickly and easily. Using the example of ophthalmic robots, this means we need to clarify how the system must be positioned and set up in relation to patients - if possible, in such a way that manual readjustment is no longer necessary.

Could AI also help to improve the interaction between surgeons, patients, and the robotic system?

Mathis-Ullrich: An operation is often difficult from an ergonomic point of view, because surgeons and assistants work together in a very confined space with different instruments. First of all, robot cognition is needed: the system must grasp and understand the situation. To do this, we use various machine learning methods. Also, to achieve the best robot movements or actions in a given situation, we explore methods of AI. Since a lot of data is needed to learn movements, we train our systems in simulation before evaluating the methods in the real world.

We are not trying to develop completely autonomous robotic systems. Rather, it's about networking intelligent systems to make certain jobs easier. Our subject area is very interdisciplinary. Medical and technical professions work together here, which is why we try to bring students from both disciplines together at a very early stage.