TH-A-9A-09: Optimal Feedback Control with Feed Forward for a Robotic 4D Stage in Frameless Stereotactic Radiosurgery




A novel 4D robotic system was developed to correct both translational and rotational head motion deviations during frameless image-guided stereotactic radiosurgery (SRS) procedures. The work presents a systematic approach to optimize correction performance.


The system included a robotic 4D (XYZ+Pitch) stage, a head position sensing unit and a control computer. The measured head position signal was processed and a resulting command was sent to the interface of a 4-axis motor controller, through which 4 stepper motors were driven to perform motion compensation. Due to the displacement between the lesion and the pivot of the 4D stage, the rotation of the 4D stage leads to both intended rotational and undesired translational motion of the lesion. To address the issue, feedback control with feed forward was utilized in control law design, and coupling coefficients were online tuned to improve control performance. The synchronization of all 4 axes motion was found to be crucial in the decoupling control.


A dynamic 6D head phantom was placed on the top of the 4D stage and was used both to perform pure 4D displacements together with simulating actual framless SRS patient head motion. The 4D stage was applied to compensate such motion. The translational displacement 7 mm and pitch displacement 0.8 degree were corrected in 5 seconds without overshoot, and the steady errors were less than 0.15 mm and 0.04 degree, respectively. For a 15 minutes test with 0.25 mm and 0.07 degree target position tolerance, a subject within tolerance in 1% of the time without motion correction, while this was increased to 99.5% with motion correction.


We report a general control design approach for both translational and rotational head motion correction. The experiments demonstrated that control performance of the 4D robotic stage meets real time sub-millimeter accuracy requirement of frameless SRS.

Research was partially funded by NIH Grant T32 EB002103-21 from NIBIB. Contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIBIB or NIH.