The Region-based Virtual Fixtures Control for Invasive Avoidance in Hands-on Robot-assisted Mandibular Angle Split Osteotomy

In robot-assisted surgery, osteotomy is a common task with surgical tools such as reciprocating saws, the miller, and the ultrasound osteotome. Although, the mechanism of different surgical tools obtains different kinematics constraints. The restriction in the invasive direction is one of the most challenging issues in the osteotomy scenarios as it directly determines the performance for surgical safety. Shared control techniques based on Virtual Fixtures (VFs) can be an effective way to help the surgeon prevent overshoot during operation. This letter presents a region-based virtual fixture to restrict the robot's motion in the reference directions. In order to generate a task plane and divide the forbidden area, a PCA-algorithm-based method is proposed with the selected several points from the medical image-based software. A variable admittance policy based on the distance in the invasive direction is adopted to adjust the admittance scale for the hands-on robot system. As a result, the surgeon will obtain different robotic velocities in different areas. An accuracy analysis is applied to the error between the task plane and the manually selected points. The maximum projection error is 1.07 mm. The motion analysis in different areas is given, and the minimum distance in the invasive direction is 0.2229 mm to the border between the marginal area and the forbidden area.