Body follows head: Bridging the gap between rigid and flexible endoscopes

A successful endoscopic intervention hinges on two key factors: the endoscope must have excellent maneuverability and high structural rigidity. Unfortunately, today's devices cannot meet both requirements at the same time. Tim-Lukas Habich wants to change that by bridging the gap between flexible and rigid robots.

In this MEDICA.de interview, he explains his research on a hydraulic snake-like robot for endoscopy.

Mr. Habich, you are developing a robotic system as part of the HALEy research project. Why is there no such device in medicine already?

Tim-Lukas Habich: The project name HALEy stands for "Hydraulic snake-like robot for endoscopy". It is a project of the Deutsche Forschungsgemeinschaft (German Research Foundation, DFG) that supports basic research and entails our collaboration with the Heidelberg University.

Conventional endoscopes are either fully flexible or fully rigid. Rigid endoscopes maintain high stiffness or rigidness but cannot reach target areas that are hidden or difficult to reach. For example, a gastroscopy requires a physician to look at the inside of the stomach. It would be impossible to accomplish this via your esophagus with a rigid endoscope. That is why the physician uses a flexible endoscope to warrant excellent maneuverability. Except for the actuated tip, it has a purely passive structure. The rigidity at the distal end is too low if you want to manipulate tissue with it and provide the doctor with a stable working platform.

Or let’s move past medicine and look at technical applications. Take a borescope, used to inspect aircraft engines, for example. The instrument allows you to check turbine rotor blades or the compressor section. This is where flexible endoscopes come into play for viewing along curvilinear paths. It’s easy to maneuver them into the difficult-to-reach target area. However, the device is too flexible to be used to do maintenance work on the blades. 

And this is where your research wants to bridge the gap?

Habich: The core benefit of snake-like manipulators or robots is that the passive shaft is replaced by many actuated segments. However, you rarely see stiffenable mechanisms – even in research. As part of the HALEy project, we examine the principle of the hyper-redundant robot, which features discrete joints. Instead of a passive structure with an actuated tip, you have a snake robot whose body can be fully controlled. This provides the desired rigidity at the end effector to manipulate tissue or work on machines. Cable-driven continuum robots also show great promise in this setting. Our colleagues study the latter for the aforementioned engine maintenance application.

You added something new to the HALEy robot. What is it?

Habich: Fluidic actuators with variable stiffness. Two antagonistic bellows arranged around a tilt axis facilitate the continuous adjustment of both the tilt angle and the joint stiffness. To envision this, think of a human arm where we can change the angle between forearm and upper arm continuously. If you flex the biceps and triceps at the same time, you can modulate the stiffness at a constant hand position.

If you stack several actuators on top of each other - as is typical for snake robots - the structure is stiffened while the robot configuration remains constant. This combines the benefits of rigid and flexible endoscopes: the robot can be maneuvered along curvilinear anatomical paths and can stiffen itself in the target area to create a stable working platform.

A previous project at our Institute focused on a metal snake robot that was actuated electromagnetically. It is not suited for medical CT or MRI uses as part of a biopsy, for example. The actuators heated up and were binary, meaning they could only tilt left or right. Our 3D-printed fluidic actuators are much better suited for medical applications: They do not interfere with the devices; they do not heat up and they facilitate continuous adjustment. The actuators are also made of solid and soft material. This makes the robot intrinsically flexible when it comes into direct contact with the patient or physician.

How could surgeons use your endoscope?

Habich: Last year, I focused on the telemanipulation of snake robots, that being the strategy of how doctors can intuitively control the robot remotely. Like real snakes, snake-like manipulators exhibit many degrees of freedom in movement. Humans can only specify a maximum of six degrees of freedom: three translational motions and three rotational motions of the hand. Some haptic input devices on the market identify the six-dimensional pose of the hand during surgery and provide force feedback between the robot and the tissue. It means the surgeon feels what the robot is “feeling”. The old computer game “Snake” uses the follow-the-leader-principle. Our snake robot should do the same, meaning the entire body follows the lead of the snake’s head. This way, it will have the ability to follow curvilinear anatomical paths.

What's your next big goal for HALEy?

Tim-Lukas Habich: Our research project started in October 2020 and runs for two and a half years, which means it is nearing the end. Right now, we are considering our next steps. It would be interesting to find a suitable industry partner with expertise in microsystems technology for the miniaturization of fluidic actuators. If two university institutes join forces and conduct basic research, you make headway in the right direction. Our soft actuator control strategies are an example of this. I am looking into AI approaches in this setting as the soft material is difficult to characterize and control with conventional model-based methods. The follow-up project will focus on this aspect.

But, like I said, if we could join forces with an industry partner who specializes in miniaturization, we could likely build a snake robot with multiple actuators yet a small diameter. Our current snake robot is still very large and a long way away from the dimensions of a flexible endoscope.