26.2.1 ROSA Spine

The ROSA Spine device is a robotic arm with tracking capabilities and a navigation system. Its platform includes the robotic arm attached to a base coupled with a haptic sensor and a touch screen surgical workstation as well as an optical navigation camera.

The patient is positioned prone on a radiolucent operating table and the surgical site is prepped in sterile fashion. The O-arm (Medtronic) is covered with a sterile drape and positioned opposite the surgeon and slightly to his or her left in order to obtain intraoperative imaging. The ROSA Spine robot is covered with a sterile drape and positioned adjacent to the table and with the base perpendicular to the patient. The optical navigation camera is positioned at the foot of the bed and angled to face the operative site. ^{4 ,​ 5}

A percutaneous reference pin is placed at the iliac wing. The robot’s position is then coregistered to the patient reference using the optical camera. Next, the fiducial box attached to the robot arm is positioned using the robot’s haptic properties. The fiducial box should be placed just over the skin at the operating site. CT imaging is obtained using the O-arm in breath-hold mode and subsequently transferred to the ROSA Spine surgical workstation. The utilization of breath-hold mode allows the robot to track movements of the spine induced by respiration. Recording of 3D images is conducted with automatic recognition by the fiducial box. ^{4 ,​ 5}

Following image registration, the surgeon begins surgical planning. The first step is planning the trajectory for bilateral pedicle screw placement at the targeted levels to be fused. The surgeon choses his or her desired entry point and target point for the trajectory of each screw and then determines the length and diameter of each screw. During the planning process, the surgeon has access to the axial, sagittal, and coronal views on the robot’s workstation and can also look along or perpendicular to the planned trajectory.

The robot arm is positioned along the trajectory and the movement tracking function is activated. Movement of the patient’s body is tracked and accounted for with the robot’s real-time navigation capability which allows instruments to remain along the planned trajectory throughout the procedure. Dilators are attached to the robot arm and positioned just above the skin at the first level to be instrumented. A small skin incision is made and the dilators are placed through the skin and muscle to access the entry point on the pedicle. A drill with a 3-mm bit is inserted through the dilator and an entry hole is drilled into the pedicle. The drill bit and all other instruments are tracked using real-time navigation. It is recommended to drill 20 mm through the bone and perform back-and-forth drill movements until no resistance is met in order to avoid a “ripping effect” between the drill and the bone. The back-and-forth movement serves to widen the entry point. ⁵

A guide tube needle is placed through the dilator and then through the pedicle into the posterior vertebral body and a guide wire is inserted. The guide wire and all instruments are tracked with computer-aided navigation utilizing one of the percutaneous ancillary systems compatible with the ROSA Spine (Sextant, Longitude, or Socore) and monitored in real time. If the position of an instrument displayed by the navigation system differs from the planned trajectory, the true position can be confirmed using intraoperative fluoroscopy. The guide-tube needle is removed and dilators are placed through the muscles over the guide wire. The pedicles are then tapped and screws are placed percutaneously via the guide wire. The two percutaneous incisions are then joined in order to place a minimally invasive retractor and expose the articular facet. Next, a unilateral facetectomy is performed and the nerve root is released and retracted. TLIF may be performed according to each surgeon’s preference. Thorough diskectomy is performed and ultimately a TLIF cage is placed with the option to use real-time navigation during diskectomy and cage insertion to minimize use of fluoroscopy. ^{4 ,​ 5}

The procedure is conducted in a similar fashion for an open approach. The steps for intraoperative imaging and pedicle screw planning are the same as for a minimally invasive approach. When the lamina and articular facets of the vertebral bodies to be fused are exposed, the robot arm is positioned and the pedicles are drilled with robotic assistance. Pedicle screws are placed similarly with robot assistance.

Following arthrodesis, a CT scan is performed to verify final positioning. Once acceptable placement of hardware has been confirmed, the wound is closed in standard fashion.

26.2.2 SpineAssist Robot

The SpineAssist robot is a parallel manipulator device that utilizes a semiactive mode in positioning surgical instruments. The main components of the system are the robot itself and the SpineAssist workstation. The robot uses three outrigger arms which are designed to accommodate a drill guide sleeve. There are also two mounting options for the robot: the spinous process clamp and the Hover-T Minimally Invasive Frame (MAZOR). The spinous process clamp is used to attach the robot directly to a single spinous process near the target levels through a small incision. The spinous process clamp allows access to two motion segments with the robot attached to a specially designed bridge that has three accommodations for the robot. The Hover-T Minimally Invasive Frame is used to secure the robot to a frame attached to the patient with a K-wire and two Steinmann pins. The frame has a central bar that is aligned along the spine with a base that allows the robot to be attached in one of 19 different positions. ⁶

The first step in utilizing the SpineAssist platform involves preoperative imaging and planning. A CT scan is obtained using the SpineAssist protocol. The CT must use 0.4- to 1-mm slices all in parallel and all in the same dimensions without compression. The images are then transferred to the SpineAssist workstation or to the surgeon’s personal computer. The SpineAssist software is then able to generate 3D virtual images of each vertebra of interest to be used for surgical planning. The surgeon uses the reconstructed 3D images to plan the optimal entry point, trajectory, length, and diameter of pedicle screws. If the surgeon uses a computer other than the SpineAssist workstation, that device must be present in the operating room or the plan must be transferred via portable storage to the SpineAssist workstation. ⁶

Several steps must take place in the operating room prior to the initiation of the procedure, and these should ideally occur prior to the patient being in the room. First, the accuracy of the SpineAssist system must be verified. The device is mounted on a jig with three-hole positions that are recognized by the software. The verification process is initiated and the software positions the robot according to the holes in the jig. A K-wire is placed through the drill sleeve and into one hole at a time to verify the starting point and trajectory. ⁶

Another important step that can take place at the same time as system accuracy verification is the calibration of intraoperative fluoroscopy. The system utilizes a specifically designed phantom that is attached onto the image intensifier of C-arm. The phantom has two surfaces that contain metal beads which are recognized by the software and used to calibrate C-arm images and control for distortions caused by surrounding electronic fields in the operating room. Once the phantom is attached, the C-arm is aligned with the orientation of the patient and an anteroposterior (AP) and lateral X-ray without any objects in the field of view for calibration. Four additional images are required once the patient is in position to complete calibration. ⁶

The patient is positioned prone on a radiolucent operating table and prepped and draped in sterile fashion. At this point, either the spinous process clamp or the Hover-T Minimally Invasive Frame is attached to the patient. To affix the spinous process clamp, the target level is identified and an approximately 4-cm incision is made over the spinous process. The clamp is attached to the spinous process and the bridge is attached to the clamp. This can be done with or without fluoroscopic guidance. To attach the Hover-T Minimally Invasive Frame, a K-wire is placed through the frame into one of the spinous processes and two Steinmann pins are then placed through the frame into the posterior-superior iliac spine (PSIS) on either side. Pin placement can be performed with or without fluoroscopic guidance. ⁶

Once the preferred mounting platform is anchored to the patient, four X-rays are taken to complete the image registration process. First, an AP and lateral X-rays are taken without the targeting device in place. Then the targeting device is attached to the mounting platform and another set of AP and lateral X-rays is taken. At this point, the image registration process is complete and no further imaging is required throughout the operation. The software automatically registers the four intraoperative images to the preoperative CT images and planned screw trajectories. The surgeon then uses the workstation to verify that the synchronized images are matched appropriately (the software also independently evaluates the accuracy of the image match). Imaging should be repeated if the fluoroscopic images are not well matched to the preoperative CT. The surgeon’s plan is displayed as an overlay on the synchronized images once accurate image registration is achieved. ⁶

The SpineAssist system is now ready for pedicle screw placement. The surgeon selects a vertebral body to instrument and the software determines the appropriate position to attach the SpineAssist robot based on which mounting platform is used. The software also determines which arm to use. If the Hover-T Frame is used, the software displays all combinations of robot positions and arms that the surgeon may select. The robot is secured in the chosen position and the software moves the robot to achieve the planned entry point and trajectory. Next, the arm is attached to the robot’s top plate and a cannulated drill guide is placed into the guide sleeve. A drill bit is then introduced into the cannulated drill guide. A small skin incision is made to introduce an obturator and bluntly dissect through soft tissues to bone along the trajectory of the drill. The drill is advanced along the trajectory to the entry point and a hole is made through the cortex. A K-wire is then placed in the hole and advanced into the vertebral body. Next a screw hole is drilled and the pedicle is probed to check for any breach followed by insertion of the pedicle screw. This process is repeated at each level until all screws are placed. ⁶
Once pedicle screws are in place, dilators can be placed in order to perform a facetectomy. The nerve root is released and retracted and the surgeon conducts a thorough diskectomy. TLIF cages are then placed and positioning is confirmed with fluoroscopy.