In their work to build semi-autonomous colonoscopy, researchers have made a breakthrough, using a robot to direct a medical device into the body.
The milestone brings closer the prospect of an intelligent robotic device being able to direct tools to biopsies in specific places in the body or enable the inspection of internal tissues.
A doctor or nurse will still be on hand to make clinical decisions, but a robotic system offloads the daunting task of manipulating the unit.
‘Enabling the future of colonoscopy with intelligent and autonomous magnetic stimulation’ is the result of 12 years of study by the University of Leeds’ international team of scientists.
The research is being published in the scientific journal Nature Machine Intelligence today (Monday, 12 October).
Next year or in early 2022, patient trials using the device could begin.
The research is being supervised by Pietro Valdastri, Professor of Robotics and Autonomous Systems at Leeds. “He said:” Colonoscopy offers physicians a window deep into the human body into the environment and it plays a crucial role in the screening of diseases such as colorectal cancer. But for decades, the technology has remained largely unchanged.
What we’ve built is a device that is easier to run for physicians or nurses and less frustrating for patients. It is an important step in making colonoscopy even more widely accessible — essential if colorectal cancer is to be diagnosed early.
Since the machine is simpler to use, scientists hope that this will increase the number of providers who can perform the procedure and allow colonoscopy to be more available to patients.
In order to analyze the rectum and colon, a colonoscopy is a procedure. Conventional colonoscopy is done using a semi-flexible tube inserted into the anus, a procedure that some patients find so painful that an anaesthetic is required.
Flexible Colonoscope Magnetic
A smaller, capsule-shaped device has been developed by the research team, which is connected to a narrow cable and inserted into the anus and then directed into place, not by the doctor or nurse moving the colonoscope, but by a magnet placed over the patient on a robotic arm.
The robotic arm swings as it manoeuvres the capsule around the patient. The theory that magnetic forces attract and repel is the foundation of the method.
In the capsule inside the body, the magnet on the outside of the patient interacts with tiny magnets that navigate it through the colon. It would be less painful than traditional colonoscopy, the researchers add,
It is possible to direct the robotic arm manually, but it is a technique that is difficult to learn. In response, different levels of robotic assistance were created by the researchers. This latest study examined the efficacy of the various levels of robotic assistance in assisting non-specialist workers to carry out the operation.
Robotic assistance levels
Direct control over robots. This is where, via a joystick, the operator has direct control of the robot. There is no help in this situation.
Intelligent Teleoperation with Endoscopes. The operator focuses on where they want the capsule in the colon to be placed, leaving the robotic system to measure the robotic arm movements needed to position the capsule in place.
Navigation semi-autonomously. Using computer vision, the robotic device autonomously navigates the capsule through the colon — although the operator can override this.
10 non-expert workers were asked to get the capsule to a point during a laboratory simulation.
10 non-expert workers were asked to get the capsule to a point inside the colon inside 20 minutes during a laboratory simulation. They have done it five times, using three different levels of support.
The participants had a 58 per cent success rate using direct robot control. Using intelligent endoscope teleoperation, that increased to 96 percent, and 100 percent using semi-autonomous navigation.
Two participants were asked to maneuver a traditional colonoscope into the colon of two anaesthetized pigs at the next stage of the experiment, and then to repeat the process with the magnet-controlled robotic device using various levels of assistance. To ensure that the animals were not injured, a vet was in attendance.
Participants were graded both physically and mentally on the NASA Task Load Index, a measure of how demanding a task was.
Source of Story: Materials from Leeds University. Note: For style and length, material can be edited.