Principles of Image-Guided Spinal Navigation

Image-guided spinal navigation is a computer-based surgical technology that was developed to improve intraoperative orientation to unexposed anatomy during complex spinal procedures.1,2 It evolved from the principles of stereotaxy that have been used by neurosurgeons for several decades to help localize intracranial lesions. Stereotaxy is defined as the localization of a specific point in space using 3D coordinates. The application of stereotaxy to intracranial surgery initially involved the use of an external frame attached to the patient′s head. However, the evolution of computer-based technologies has eliminated the need for a frame and has allowed for the expansion of stereotactic technology into spinal surgery.

The management of complex spinal disorders has been greatly influenced by the increased acceptance and use of spinal instrumentation as well as the development of complex operative exposures. Many of these techniques require the surgeon to have a precise orientation to spinal anatomy that is not exposed and, therefore, not directly visualized in the surgical field. With the proper surgical exposure, direct visualization of the rostral, caudal, medial, and lateral aspects of the exposed bony anatomy is straightforward. However, it is the imprecise estimation of structures deep to the exposed anatomy that potentially confounds surgeons and leads to misplaced instrumentation and possible neurological or vascular injury. In particular, the various fixation techniques that require placing screws into the lateral masses of C1 and across the C1-C2 joint space require “see-through visualization” of the unexposed spinal anatomy.

Although conventional intraoperative imaging techniques such as fluoroscopy are helpful, they are limited in that they provide only two-dimensional imaging of a complex 3D structure. Consequently, the surgeon is required to extrapolate the third dimension based on an interpretation of the images as well as the surgeon′s knowledge of spinal anatomy. This “dead reckoning” of anatomy can result in varying degrees of inaccuracy when placing screws into the unexposed spinal column.3–6

The main objective of computer-based image guidance is to establish a spatial relationship between image data of the surgical field and its corresponding intraoperative anatomy. The image data typically used for navigation purposes can be a preoperatively or intraoperatively acquired CT scan or an intraoperatively acquired fluoroscopic image. A spatial relationship can be created between the image data and the corresponding surgical anatomy because each represents a series of points, or coordinates, in a 3D space. Each point can be defined by a specific x, y, and z Cartesian coordinate. With defined mathematical algorithms, a specific point in the image dataset can be matched to its corresponding point in the surgical field. After several of these points are matched, or registered, it becomes possible to select random points in the surgical field and have the corresponding points in the image dataset displayed. These image data points can then be reformatted into multiplanar images of the surgical field, greatly enhancing the surgeon′s intraoperative visualization of the pertinent surgical anatomy.

CT-based image-guided spinal navigation was developed to provide real-time axial imaging of selected screw trajectories. It allows for the intraoperative manipulation of multiplanar CT images that can be oriented to a selected point in the surgical field and helps minimize much of the guess work associated with complex spinal surgery. Although it is not an intraoperative imaging device, it provides the spinal surgeon with superior image data compared with conventional intraoperative fluoroscopy. It improves the speed, accuracy, and precision of complex spinal surgery while, in most cases, eliminating the need for an intraoperative fluoroscopy.

A variety of navigational systems have evolved over the past decade. Although each of these systems may differ slightly, the common components of these systems include an image-processing computer workstation interfaced with a two-camera optical localizer ( Fig. 38.1 ). When positioned during surgery, the optical localizer emits infrared light toward the operative field. A handheld navigational probe mounted with a fixed array of passive reflective spheres serves as the link between the surgeon and the computer workstation ( Fig. 38.2 ). Alternatively, passive reflectors may be attached to standard surgical instruments.

The spacing and positioning of the passive reflectors on each navigational probe or customized trackable surgical instrument are known by the computer workstation. The infrared light transmitted toward the operative field is reflected back to the optical localizer by the passive reflectors. The information is then relayed to the computer workstation, which can then calculate the precise location of the instrument tip in the surgical field as well as the location of the anatomic point on which the instrument tip is resting. Multiplanar images through the selected point are generated, providing the surgeon with optimal spatial information for performing the surgical plan.

Spinal anatomy itself can be used as a frame of reference to establish the spatial relationship between image data and surgical anatomy. Accordingly, neither a stereotactic frame nor surface-mounted fiducials are necessary because bone landmarks on the exposed surface of the spinal column provide the points of reference necessary for image-guided navigation. Specifically, any anatomic landmark that can be identified intraoperatively as well as in the preoperative image dataset can be used as a reference point. The tip of a spinous or transverse process, a facet joint, or a prominent osteophyte can serve as potential reference points ( Fig. 38.3 ). Because each vertebra is a fixed and rigid body, the spatial relationship of the selected registration points to the vertebral anatomy at a single spinal level is not affected by changes in body position.

The initial step of spinal navigation is the creation of a spatial relationship between the image data and the corresponding surgical anatomy. This step is called registration. Three different registration techniques can be used for spinal navigation: paired point registration, surface matching, and automated registration. Paired point registration involves selecting a series of points in a CT dataset and matching them to their corresponding points in the exposed spinal anatomy. The registration process is performed immediately after surgical exposure and prior to any planned decompressive procedure, allowing for the use of the spinous processes as registration points.

A specific registration point in the CT image dataset is selected by highlighting it with the computer cursor. The tip of the probe is then placed on the corresponding point in the surgical field, and the reflective spheres on the probe handle are aimed toward the camera. Infrared light from the camera is reflected back, allowing the spatial position of the probe′s tip to be identified. This initial step of the registration process effectively matches the point selected in the image data with the point selected in the surgical field. When a minimum of three such points are registered, the probe can be placed on any other point in the surgical field and the corresponding point in the image dataset will be identified on the computer workstation. Additionally, multiple planar images centered on the selected point will be displayed.

Alternatively, a second registration technique called surface matching can be used. This technique involves the selection of multiple, nondiscreet points on the exposed and debrided surface of the spine in the surgical field. This technique does not require a prior selection of points in the image set though several discreet points in both the image dataset and surgical field are frequently required to improve the accuracy of surface mapping. The positional information of these points is transferred to the workstation, and a topographic map of the selected anatomy is created and matched to the patient′s image set.

A third registration technique is termed automated registration. This technique is used with navigation, employing intraoperatively acquired fluoroscopic or CT images. During image acquisition, a reference grid internal to the imaging device is interposed between the image source and the spinal anatomy. Registration is then performed by the imaging system itself without a need for input from the surgeon. Up to five spinal levels can be registered with this technique without compromising accuracy.

The purpose of the registration process is to establish a precise spatial relationship between the image space of the data with the physical space of the patient′s corresponding surgical anatomy. If the patient is moved after registration, this spatial relationship is distorted, making the navigational information inaccurate. This problem can be minimized by the optional use of a spinal tracking device that consists of a separate set of passive reflectors mounted on an instrument that can be attached to the exposed spinal anatomy. The position of the reference frame can be tracked by the camera system. Movement of the frame alerts the navigational system to inadvertent movement of the spine. The system can then make correctional steps to keep the registration process accurate and eliminate the need to repeat the registration process.

When CT-based navigation is used, registration is performed and the navigational probe is placed on a surface point on the registered vertebrae. When activated, the system will immediately display three separate reformatted CT images centered on the corresponding point in the image dataset. Each reformatted image is referenced to the long axis of the probe. If the probe is placed on the spinal anatomy directly perpendicular to its long axis, the three images will be in the sagittal, coronal, and axial planes.

A trajectory line representing the orientation of the long axis of the probe will overlay the sagittal and axial planes. A cursor representing a cross-section through the selected trajectory will overlay the coronal plane. The insertional depth of the trajectory can be adjusted to correspond to selected screw lengths. As the depth is adjusted, the specific coronal plane will adjust accordingly with the position of the cursor demonstrating the final position of the tip of a screw placed at that depth along the selected trajectory.

As the probe is moved to another point in the surgical field, the reformatted images as well as the position of the cursor and trajectory line will change. The planar orientation of the three reformatted images will also change as the probe′s angle relative to the spinal axis changes. When the probe′s orientation is not perpendicular to the long axis of the spine, the images displayed will be in oblique or orthogonal planes. Regardless of the probe′s orientation, the navigational workstation will provide the surgeon with a greater degree of anatomic information than can be provided by any intraoperative imaging technique.