Surgical Robot Assistant



          (WSU): Daniel Whitman, Lyubomir Zagorchev, Ardeshir Goshtasby

          (WKNI): Martin Satter, Teodore Bernstein

          (Motoman): Greg Webb, Erik Nieves, Joe Lill

Sponsors: WKNI (Wallace Kettering Neuroscience Institute); Navigation system contributed by Medtronic; Robot contributed by Motoman


Pedicle screw fixation is an established medical procedure for correcting congenital and acquired spinal deformities. The main challenge in this procedure is to ensure that the screws remain in the main axis of the pedicle. Misplaced pedicle screw complications are mostly neurological, although screws that penetrate the anterior cortex of the vertebral body can cause vascular and visceral damage. The rate of misplacing pedicle screws in the lumbar spine varies from a few percent with a navigation system to 40% without navigation.

Accurate fixation of the screws requires insertion in the vertebral body through the axis of the pedicle. Therefore, during placement, the exact location of the pedicle axis is required. When vertebrae do not have a normal shape for pathologic reasons, use of a-priori rules such as a prespecified angle between the pedicle and the spineous process is not reliable. Therefore, prior to intervention, a trajectory to define the exact location of the pedicle axis is defined on a CT or MR image. In the operating room, the trajectory is reproduced using intra-operative images and a guidance system. With open surgery, the entry point of a surgical tool can easily be found since there is direct access to it, but the direction of insertion to the bone is very difficult without registration of a pre-operative image with intra-operatively obtained images.

Using an image-guided surgery system an intervention can be made less invasive, performing pedicle screw placement percutaneously. A less invasive operation reduces complications and rehabilitation time. It has been shown that image-guided surgery of the lumbar spine improves the accuracy of in-vivo pedicle screw placement. To enable accurate insertion of pedicle screws in vertebral bodies, a steady arm is required that will keep a fixed drilling angle. This is achieved using the steady arm of a robot. The robot arm will assist the surgeon to drill a hole in a vertebral body in the exact desired position and orientation.

In a system being developed by Wright State University, Wallace-Kettering Neuroscience Institute, and Motoman, a surgical navigation system and a robot arm are operated simultaneously and in the coordinate system of the patient. While surgery is being performed, the motion of the patient is tracked by an infrared stereo sensor and fed to both the robot and the navigation system. The surgeon by viewing an MR or CT image of the patient that is registered with the patient, and a robot arm that works in the coordinate system of the patient and thus in the coordinate system of the image, controls the robot arm to move to the desired position with respect to the patient and aid the surgeon to drill a hole in a vertebral body in the planned position and orientation. The organization of this system is shown below.

The stereo infrared sensor is shown on the right side. The model spine is shown in the middle. A pointer fixed to the patient with four infrared reflectors in fixed relations of each other attached to it define the coordinate system of the patient. Similar reflectors in fixed relations of each other are attached to the robot arm, making it possible for the infrared sensor to locate it with respect to the patient in real-time. By registering the patient with a CT or MR image of the patient, the relation between image coordinates and patient coordinates is determined, keeping track of the patient and robot simultaneously and mapping their positions to the MR or CT image space where surgery plan is made.

The same setup from the front view. The big screen in the background shows the registered patient and CT or MR image.

The Medtronic navigation system is shown in the left. Image, patient, and the instrument held by the robot are overlaid and shown in the same window. Motion of the patient and the robot are tracked by the sensor and mapped to the image for viewing in real-time.

The robot arm from close up. The reflectors attached to the patient and to the robot arm are used to locate and map the patient and the instrument held by the robot arm to the image space.

And, these are the folks who are doing the actual work.


For more information contact A. Goshtasby (