INTRODUCTION In the context of robotic assistance to medical gestures, we propose solutions to stabilize the ultrasound (US) image by actively compensating for the physiological motions of the patient. The considered applications are for instance the assistance for diagnoses or hepatic tumor biopsies where the liver and the tumors mainly undergo the respiratory motion. Other clinic applications, such as prostate cancer brachytherapy have been identified in [1] that could benefit from such robotic image stabilization. The per-operative image provided by the US probe and the contact force applied to the probe are used to control the six degrees of freedom (dof) of the robot. To deal with the low quality of the US images, we propose to use the intensity of the image as visual features in a visual servoing control loop [2][3]. This vision control is associated with a force control to ensure a constant force applied to the probe. Finally, a predictive controller is implemented in the control law to take advantage of the repetitiveness of the physiological motions. We present first ex-vivo robotic results on animal tissues where we compensate for the 3D motions using successively 2D and 3D probes.
[1]
Alexandre Krupa,et al.
Intensity-based direct visual servoing of an ultrasound probe
,
2011,
2011 IEEE International Conference on Robotics and Automation.
[2]
Gregory D. Hager,et al.
Real-time Motion Stabilization with B-mode Ultrasound Using Image Speckle Information and Visual Servoing
,
2009,
Int. J. Robotics Res..
[3]
Alexandre Krupa,et al.
Intensity-based visual servoing for non-rigid motion compensation of soft tissue structures due to physiological motion using 4D ultrasound
,
2011,
2011 IEEE/RSJ International Conference on Intelligent Robots and Systems.
[4]
Luc Soler,et al.
Model predictive control for compensation of cyclic organ motions in teleoperated laparoscopic surgery
,
2006,
IEEE Transactions on Control Systems Technology.