Simulation Model which can Visualize the Shape of the Colonoscope using Orientation Sensor Network

Jaewoo Lee, Genya Ukawa, Shuna Doho, Zhuohua Lin, Hiroyuki Ishii, and Atsuo Takanishi


orientation sensor network, kinematic chain, shape, smoothness


The Simulation Model which can visualize the shape of commercial colonoscope is proposed. This model uses kinematic chain which approximates the shape of colonoscope by serially connected multiple joint-link pairs. As the input to the model, orientation information is used. Orientation information is extracted from the Sensor Network. The sensor network consists of connecting a number of sensor units with CAN bus. Each sensor unit also has two kinds of sensor: triad of accelerometer and triad of magnetometer. Roll and pitch angle are estimated based on the accelerometer signals. Yaw angle is estimated based on the magnetometer signals. By using kinematic chain model and orientation as input, we estimate the key points which are positions of sensor units on the colonoscope. Then, we implement spline interpolation to find the intermediate points which constitutes the shape of colonoscope. We made experiment in order to check whether model has enough accuracy. Firstly, we compared estimated curve with sine curve which is used as a ground truth. Secondly, we investigated the effect of number of sensor on the accuracy. Final result says that it has enough estimation ability.

Important Links:

Go Back