Surgeons can now navigate through body spaces more precisely and safely
Traditionally, surgeons working on deep anatomical regions need to disrupt tissue with invasive incisions. Intraluminal endoscopic surgeries are a revolutionary solution, utilising natural body lumens such as the colon, a ureter or the vasculature as pathways. Compared to open surgeries, intraluminal procedures offer numerous advantages, including faster recovery times and reduced scarring. Although less invasive, this approach demands high surgical skill and dexterity, as these lumens are fragile and bend in many directions. Additionally, visualisation is often limited, creating more challenges. Partly funded by the Marie Skłodowska-Curie Actions programme, the ATLAS project has made major contributions to solving these issues by integrating robotic automation. Advancements include 3D printing developments, sensor integrations and use of AI algorithms.
3D printing and intraluminal technology
Project coordinator Emmanuel Vander Poorten of host institution KU Leuven states: “The project demonstrated the feasibility of using low-cost 3D printers to produce (print) steerable instruments. Embodied with artificial muscles, these 3D-printed structures can be introduced in body lumens and smoothly navigate to deeply seated targets.” The project team developed deep learning AI-methods to control these complex instruments. ATLAS researchers also produced structures that bend into complex curves when an external magnetic field is applied. Magnetic actuation is advantageous for, among other things, guidewire applications. Instead of employing multiple pre-bend guidewires, the clinician can steer and adjust the angles of a single wire via magnetic actuation, therefore increasing efficiency.
Safe sensor technology and AI solutions to surgical challenges
Project outcomes offer a solution to difficult problems like precisely locating the exact position of the catheters. Instead of traditional fluoroscopy, non-radiative approaches for example, based on use of optical fibres are developed. The reduced radiation lowers the risk to the patient and clinicians while also enabling a 3D view. Current challenges of AI in surgery concern the amount of input data needed – within lumens it is impossible to calculate all possibilities. ATLAS innovated a solution based on anomaly detection, overcoming this challenge. Vander Poorten adds: “Building upon advances in deep learning, AI algorithms were developed that operate in real time and have the potential to offer clinicians unprecedented insights on the tissue from within the lumen.” With these insights, surgeons can navigate more safely through the fragile passages.
Robotic surgery’s future is in the hands of young researchers
Following the Horizon 2020 programme Fostering new skills by means of excellent initial training of researchers, ATLAS provided a unique, multidisciplinary academic programme for early-career researchers interested in robotics at the interface of intraluminal surgery. The project’s offering is unique, as current training programmes from other institutions have narrower foci, only on single topics. Robotics is a highly multidisciplinary field that encompasses mechanics, electronics, sensing, software and control. By training researchers to be competent in a broader spectrum of topics, not only are they equipped to find better jobs, they can approach challenges in the field with a more informed perspective. While some young ATLAS researchers have advanced to postdoctoral positions or into the medical industry, many are still finalising their PhD research. Regardless of the stage, the potential is exciting: they are keen to apply ATLAS technology to other domains such as gynaecology and to further develop the technology within the cardiovascular, gastrointestinal and urological fields.
Keywords
ATLAS, 3D printing, AI algorithms, intraluminal surgery, deep learning, robotics, automation
