State Recognition of Pedicle Drilling With Force Sensing in a Robotic Spinal Surgical System

Abstract

Spinal orthopedic surgery is one of the typical high-risk surgeries. During the surgery, bone screws need to be inserted into the narrower vertebral pedicles and any failures in the screw-path-drilling process will directly hurt the important nerves and vessels of the patient. In this paper, a robotic spinal surgical system (RSSS) for assisting pedicle screw insertion surgery is proposed. The novel structural configuration of the RSSS can balance the gravity effects of the arms, and the workspace envelope of the RSSS can better fit the requirements of spinal surgery. The procedure of the screw insertion operation with the proposed RSSS is presented. To simulate the hand feel of the surgeon in surgical operations, a real-time force-sensing algorithm is developed for the screw-path-drilling process using a spherical or a twist drill. With the force sensing, five key states, including the initial state, the outer cortical state, the cancellous state, the transitional state, and the inner cortical state, are recognized. In particular, the clear recognition of the transitional zone between the cancellous bone and the inner cortical bone is very important as the drilled screw path ends in this zone. State recognition with force sensing can effectively improve the quality of the screw paths so as to enhance the quality of the spinal surgery. Experiments were carried out to verify the state recognition effects of the pedicle drilling process with force sensing.