Year: 2014
Type: article
Abstract: Colon cancer, the second leading cause of cancer deaths in the U.S., is curable if detected at early stages. To inspect the colon for polyps or cancerous cells, standard colonoscopy is performed by inserting a long, flexible colonoscope. Conventional colonoscopy is slow, painful, and requires sedation and insufflation. Maneuvering difficulty, possibility of loop formation, risk of tissue perforation, and a steep learning curve for surgeons are other issues with manual colonoscopy. Fusion of robotics and endoscopy is paving the way to automate this procedure to reduce trauma and discomfort to patients and facilitate wide-scale screening.Locomotion is the most challenging part of a robotic design for colonoscopy due to the intrinsic characteristics of the colon. The colon is slippery, naturally collapsed, stretchable, and loosely fixed to surrounding tissue. Many previous designs have used an inchworm approach, wheels, and/or tracks to achieve locomotion [1,2]. However, the main drawbacks of these methods are slow advancement and relatively large friction/contact with the inner wall of the intestine. Capsule endoscopy [3] is another recently developed technology. The capsule moves passively through the gastrointestinal tract without discomfort, eliminating the need for intubation, insufflation, or sedation. The main limitation of this method is inability to control the trajectory of the capsule or to localize an identified lesion. Recently, an external-magnetic-based locomotion controlled by a robotic arm has been developed [4]. This technique involves an additional expensive component (robotic arm) making the procedure less affordable, more complex for training, and longer compared with conventional colonoscopy [4]. Furthermore, due to the external magnetic field, the robotic capsule is always drawn upward to the internal wall of the intestine, creating friction and constant radial contact forces with tissues, increasing the risk of tissue damage and perforation. Here, we present a fluidic self-propelling robot that is less invasive, simple, skill-independent, easy to fabricate, inexpensive, and disposable.The robot is composed of an anal introducer, a latex tube with an inner diameter (ID) and an outer diameter (OD) of 1.6 mm and 3.2 mm, respectively, and a tip, as shown in Fig. 1. One end of the latex tube is tied over the robot's tip and the other end is connected to a poly(vinyl chloride) tube. A small portion of the latex tube is placed between the anal introducer and the tip, and the rest of it is available outside the body. The closed end of the latex tube is prestressed such that when the pressurized air is introduced into the tube, it starts to inflate first from the closed end (the tip) radially to a maximum diameter of 0.63 in. (well below the ID of the large intestine) and then axially. When the anal introducer is kept immobile, the axial expansion of the tube makes the tip of the robot advance forward. As the tip of the robot advances, more latex tube is pulled in from outside the body. This provides a swift, continuous, and long stroke. Figure 2 demonstrates the working principle of this mechanism. The balloon shortens in length by one-third as it deflates. Insofar as the descending colon accounts for almost one-third of the colon's total length, deflation of the balloon retracts the robot tip into the descending colon. Eventually, the robot can be removed by gently pulling it out without undue risk of tissue trauma.To test the functionality of the robot, a colonoscopy simulator was built (seen in Fig. 3). The simulator bed is made of soft foam, simulating the mechanical properties of the tissues surrounding the colon. A 3D channel was cut into the foam, matching the configuration of a real colon according to the anatomical length of the ascending, descending, and transverse colon as shown in Fig. 3. A commercially available synthetic colon (O-LIN-A-0005, SynDaver™ Labs, Tampa, FL), emulating the mechanical properties of living tissues, was inserted into the channel. A piece of transparent acrylic sheet was placed on top of the simulator.The tip and latex tube parts of the robot were inserted inside the synthetic colon and the anal introducer was attached to the anus and the ability of the robot to move in a straight path was tested (Fig. 4). A second experiment was conducted in the colonoscopy simulator to evaluate the capabilities of the robot in a more realistic manner. A minimum pressure of 0.138 MPa was required to get the robot to the transverse colon. Nevertheless, higher pressure might be required to overcome friction as the robot advances further. Devising a steering mechanism to ease the navigation and reach the transverse colon seemed to be necessary and is an ongoing task. In both scenarios, only the tip of the robot was in contact with the internal wall of the colon and the latex tube acted like a flexible linear actuator with minimal contact with the colon. The velocity of the robot was measured to be 86.6 and 66.5 mm/s for the straight path and the simulator, respectively, which is a 29-fold improvement over traditional colonoscopy. Robot retraction can then be performed at much lower speeds consistent with diagnostic inspection.To investigate the effect of the latex tube diameter on the performance of the robot, a latex tube with a larger diameter (ID of 3.2 mm and OD of 4.8 mm) was used. The robot was tested in the same setup as described for the smaller tube in the previous paragraph. The small tube exhibited more flexibility such that it buckled and folded inside the colon, and the large tube provided more axial force. Nevertheless, the robot with the large tube reached the same region of the colon as the robot with the small tube. In contrast to the small tube, the whole length of the large inflated tube was in contact with the internal wall of the colon.A proof of concept for a miniature robot for colonoscopy is demonstrated and presented in this paper. Our fluidic robot and novel propulsion system offer several advantages over previous designs. It took only about 3 s for the robot to propel itself from the anus to the end of the descending colon. To the best of the authors' knowledge, this is the fastest propulsion system implemented among previous robots. This will significantly reduce the time required to perform colonoscopy procedures. The friction/contact between this robot and the internal wall of the intestine is small compared with previous designs, especially those that use mechanical anchoring. This could decrease the potential risk of tissue injuries and colonic perforation that may lead to bleeding and postprocedural complications. With the ability of the developed robot to move inside the collapsed colon, insufflation of the entire colon may not be necessary as is currently used in standard colonoscopy and many other alternative robotic systems.The performance of the fluidic robot may be improved by replacing the robot's tip with a soft, gel-like material. Employing customized latex tube with a proper diameter, wall thickness, and desired elastic property could also improve flexibility and the ability to push the robot's tip forward. Integrating a small camera and a light-emitting diode to the tip of the robot for diagnostic purposes is ongoing.To reach the cecum, the robot should be able to pass other folds of the colon. Two design options are currently being explored. First, making the robot self-steerable by enhancing the robot's tip mobility. This could be achieved by increasing the rotational degrees of freedom of the robot's tip using, for instance, a ball/socket joint. Second, developing a steering mechanism similar to the system described in Ref. [4], as shown in Fig. 5, such that alternating the air pressure in three chambers would tilt the tip of the robot, easing the navigation through the large intestine.The authors gratefully acknowledge financial support from the Nebraska Research Initiative program.
Source: Journal of Medical Devices
DOI: 10.1115/1.4027076
Language: English
Subfield: Biomedical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W2053944801