Image-based hysteresis reduction for the control of flexible endoscopic instruments

The limited dexterity of conventional flexible endoscopic instruments restricts the clinical procedures that can be performed by flexible endoscopy. Advanced instruments with a higher degree of dexterity are being developed, but are difficult to control manually. Adding actuators to these instruments may make them easier to control. However, the intrinsic hysteresis that is present between the actuators and the tip of the instrument needs to be reduced in order to allow accurate control. We present an estimation algorithm that determines the hysteresis between the actuators and the instrument tip in all three degrees of freedom of the instrument: insertion, rotation, and bending. The estimation is performed on-line. The endoscopic images are used as the only feedback, and no additional sensors are placed on the instrument, which is beneficial for application in clinical practice. The estimated parameters are used to reduce the hysteresis that is present. Experimental validation showed a hysteresis reduction of 75%, 78%, and 73% for the insertion, rotation, and bending degrees of freedom, respectively.