.	CIMIL Home  |   UROBOT  |   CROBOT  |   ART  |  IMAGE | GUEST BOOK |
	.	Computer Integrated Medical Intervention Laboratory
Overview  |   Locomotion  |   Visual Control   |  Resources  |

LOCOMOTION AND STEERABLE DISTAL END

The problems of locomotion are concerned with activation systems and control sequencing of actuator elements within an environment which is slippery and has a highly varied anatomy. Steerable tips have difficulties associated with the need for remote activation and the need to negotiate tight radii without penetrating the colon wall.

EndoCrawler 2.0

Here at NTU, the CROBOT team has built various versions of EndoCrawlers. These robotic colonoscopes use pneumatic actuators to propel themselves into the tubular organ. Our first prototype, the EndoCrawler 2.0, uses balloons as actuators for propulsion using the inchworm mode of locomotion.

The EndoCrawler 2.0 Prototype

EndoCrawler 3.0

The team went on to built the EndoCrawler 3.0. This prototype has rubber bellow actuators protruding from its body. By introducing air pressure and vacuum to different bellows at different time intervals, the robot can propel itself forward. Different gaits can be implemented to suit different environments.

EndoCrawler 3.1

Experiments with the EndoCrawler 3.0 showed promising results. The CROBOT team went on to build the EndoCrawler 3.1, an improved version of the EndoCrawler 3.0.

The Entire EndoCrawler 3.1 System

The EndoCrawler 3.1 Robot

This robot carries with it an ultra-compact CCD camera, light source, surgical channel, air/water conduits and other instruments required of a real colonoscopy examination or surgery. Its distal tip is motorized and steerable to enable the robot to ?look around? and accurately position a surgical tool.

In vivo experiments were carried out using a heavily sedated pig.

Besides being able to self-propel, this lastest version of the EndoCrawler has all the capabilities of a conventional colonoscope. Experiments showed the robot's ability to traverse into a colon with the aid of a gentle pushing force from the outside. Images from the robot's CCD camera were clear enough for diagnostic purposes.

With further improvements with guided experiments, the EndoCrawler 3.1 can ease the job of the endoscopist and reduce trauma experienced by the patient. It is a viable replacement for the conventional colonoscope.

Future Works

The evolution of the EndoCrawler 3.1 involved evaluation of the performances of its predecessors through experimentations. With each robot that is built and experimented with, the research group gained more knowledge of what was required to build a robotic colonoscope. Thus, this project does not end with the EndoCrawler 3.1. Further improvements can still be made to the robot.

Size

As recommended by our consultant endoscopist, any colonoscope should preferably be smaller than 20 mm in diameter. To reduce its diameter, spaces in between the bellow actuators, in addition to the central cavity, may be used to house wires and tubings.

Steerable Distal Head

A steerable distal head is necessary in a robotic colonoscope. It will enable the endoscopist to look around the colon and to accurately position a surgical tool. However, from the experiments performed with the EndoCrawler 3.1, we realise that a flexible self conforming distal head is also required to negotiate the bends in the colon during locomotion. We can achieve this by adding a mechanical clutch to the steering mechanism. During locomotion, the clutch disengages the steering mechanism from the gearbox. This causes the distal head, which must be back-drivable, to conform to the bends in the colon. When the endoscopist wants to steer the head to a required position, the clutch will engage the driving mechanism again. Locomotion should be halted when the steering mechanism is activated since the distal head will become non compliant in the process.

Feeding Mechanism

From our experiments, it is found that a gentle external force is required to aid in the traversing motions of the robot. An automatic feeding mechanism may be developed for this purpose.

Integration of Parts

Instead of assembling separate entities, the bellow actuators and the main body of the EndoCrawler 3.1 may be cast out of a single mould. If possible, even the air/water tubings and surgical instrument channel may be moulded using the same mould cast. Some advantages arise from integration of the parts into one or fewer entities:

• The ultimate size of the robot may be further reduced.
• More of the robot can be made up of rubber, thus reducing its rigid lengths. This will enable the robot to negotiate more acute bends.
• With the integration of parts, the outer locomotive components of the robot may be mass produced and thus become disposable. After each examination, the outer sheath of the robot is detached and a new disposable sheath is re-attached. This will cut down equipment cost.
• Being disposable also means there will be no need for sterilization procedures.

Improvement of User Interface

The user interface of the system may be improved by incorporating the endoscopic view onto the same screen. Virtual distal head flipping may be possible by simply selecting the point of interest on the endoscopic view.

Colon Mapping

Using suitable positional sensors and some computation, the colon may be mapped out as a three dimensional structure during examination. This information will allow surgeons to accurately locate specific portions of the colonic wall during follow-up examinations.

Gait Study and Simulation

There is an infinite number of gait sequences that can be implemented to actuate the bellow actuators. A proper study may be carried out to determine which gait sequence would work best under different circumstances to bring about the optimal performance of the robot. This can be achieved by computer simulation of the robot in a colon.

Properties of Colonic Wall

A study may also be carried out to determine the mechanical properties of the human colon. Its complex elasticity and deformation behaviour would be useful information. With these knowledge, a more suitable material may be chosen to build the bellow actuators so that they would have a better grip on the colonic walls. Other design changes would also become more apparent and be implemented with a better understanding of the colon.

Maintained by webmaster , Last Modified: 05/28/99
Copyright © 1999, Nanyang Technological University,