Encrypted login | home

Program Information

An Image-Guided Soft Robotic Patient Positioning System for Maskless Head-And-Neck Cancer Radiotherapy: A Proof-Of-Concept Study

no image available
O Ogunmolu

O Ogunmolu1*, N Gans1 , S Jiang2 , X Gu2 , (1) The University of Texas at Dallas, Richardson, Texas, (2) UT Southwestern Medical Center, Dallas, TX


SU-E-J-12 (Sunday, July 12, 2015) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose:We propose a surface-image-guided soft robotic patient positioning system for maskless head-and-neck radiotherapy. The ultimate goal of this project is to utilize a soft robot to realize non-rigid patient positioning and real-time motion compensation. In this proof-of-concept study, we design a position-based visual servoing control system for an air-bladder-based soft robot and investigate its performance in controlling the flexion/extension cranial motion on a mannequin head phantom.

Methods:The current system consists of Microsoft Kinect depth camera, an inflatable air bladder (IAB), pressured air source, pneumatic valve actuators, custom-built current regulators, and a National Instruments myRIO microcontroller. The performance of the designed system was evaluated on a mannequin head, with a ball joint fixed below its neck to simulate torso-induced head motion along flexion/extension direction. The IAB is placed beneath the mannequin head. The Kinect camera captures images of the mannequin head, extracts the face, and measures the position of the head relative to the camera. This distance is sent to the myRIO, which runs control algorithms and sends actuation commands to the valves, inflating and deflating the IAB to induce head motion.

Results:For a step input, i.e. regulation of the head to a constant displacement, the maximum error was a 6% overshoot, which the system then reduces to 0% steady-state error. In this initial investigation, the settling time to reach the regulated position was approximately 8 seconds, with 2 seconds of delay between the command start of motion due to capacitance of the pneumatics, for a total of 10 seconds to regulate the error.

Conclusion:The surface image-guided soft robotic patient positioning system can achieve accurate mannequin head flexion/extension motion. Given this promising initial result, the extension of the current one-dimensional soft robot control to multiple IABs for non-rigid positioning control will be pursued.

Contact Email: