Intraoperative state recognition of a bone-drilling system with image-force fusion

Abstract

In pedicle screw insertion surgeries, the drilling process of the screw path is very critical to decide the success of the surgery, as the hole is drilled on a very narrow area on the vertebral pedicle. In current manual surgeries, surgeons perform operation with monitoring the medical images in navigation system and sensing operation force. To simulate these abilities, in this paper, a bone-drilling state recognition algorithm and the related system based on image-force fusion are proposed. The short-time average magnitude of thrust force, the average energy of thrust force and their gradients are used to recognize drilling state and judge whether the drilling position is appropriate. For medical image information, the preoperatively scanned medical images are combined with the real-time position information of the operation tool. And the boundary of test bone, which is used to limit the drilling motion, is found depending on the drilling direction. Fusing recognition results based on thrust force and medical images, the final recognized results are modified to be more accurate and safer to control the drilling process.