Improving the Sense of Touch for Surgical Robots
Bridging the gap between surgical simulation and surgical practice
When a knife cuts into an organ, forces push back in ways that mechanical engineers can, to some extent, predict. But other factors are also at play: Ions shift in solution within cells, causing electromechanical changes that, researchers now say, can be predicted as well. In a new model of soft tissue deformation, researchers for the first time include electromechanical changes as well as mechanical ones. The work could lead to better 3D surgical simulations and could ultimately provide surgeons at computer terminals with simulated feedback through surgical robot’s controls.
“We want to bridge the gap between surgical simulation and surgical practice,” says Yongmin Zhong, PhD, research fellow in mechanical and mechatronic engineering at the Curtin University of Technology in Perth, Australia. Zhong’s novel way of modeling soft tissue deformation was outlined in the November 2009 issue of Artificial Intelligence in Medicine.
Robots lend a helping metal hand in surgery worldwide, cutting more precisely than trembling human fingers. But the surgeons behind the joysticks cannot feel how hard to push: slicing through fatty tissue feels the same as cutting through air. When cutting by hand, “you know how hard you’re pushing, you know what damage you’re doing, “ says Julian Smith, MD, a heart surgeon at the Monash Medical Center in Melbourne, Australia, a co-author on the paper. “With robotic instruments, you get none of that.”
In previous attempts to provide a sense of touch in surgical simulations, researchers focused only on the mechanical force applied. While the mechanical force is important, Zhong explains, so are the electrical forces that come into play deeper within the tissue. For instance, charged particles like potassium swim in the plasma-like interstitial fluid between tissue cells, morphing the overall shape of the tissue.
By including the diffusion of charged particles in a set of sophisticated mathematical expressions, Zhong showed how prodding tissue shoves like-charged ions together, creating electrostatic repulsion. The model shows that this repulsion makes it harder to cut the soft tissue because it pushes back on the knife. Zhong tackled the equations with an artificial cell neural network, a much zippier problem-solver than numerical algorithms because the “cells” number-crunch as a team, instead of iteratively. It’s the computational equivalent of six people jointly solving a jigsaw puzzle instead of taking turns. Such quick computational solutions are critical in a surgery, Zhong notes, because doctors cannot work with a time lag.
“They did a very good job and it’s closer to what we can get in the real world, but it doesn’t mean the problem is solved,” says Xiaobu Yuan, PhD, associate professor in computer science at the University of Windsor. For example, poking the stomach causes it to shrink because it’s connected to the nervous system, but the new model doesn’t take that into account.
Smith plans to test if the model matches reality by putting animals under the knife. “The model is yet to be applied,” says Smith, but it has “outstanding potential.”