Chapter 20 Computed Tomography
Computed tomography (CT) is a well-established diagnostic imaging modality that has been in clinical use in medicine for several decades but only recently has become commonly used in veterinary medicine. The essential concept of a CT scanner is that an x-ray–generating tube is positioned across from a row of digital x-ray detectors in a circular gantry. The tube and detectors spin around the outside of a horse, creating a continuous radiographic image through 360 degrees. The detectors are linked to a computer system that manipulates the imaging data by a process called back projection, thereby generating cross-sectional images based on the subject density.1
CT hardware, like most imaging technology, has undergone extensive revision since its inception. One of the most important advances in CT hardware was the advent of the slip ring, allowing for helical scanning. With helical scanning, the x-ray tube and detectors continuously spin around the horse while it translates or moves through the center of the gantry. The resultant path of the x-ray tube relative to the horse is helical, giving the technique its name. Another, equally important, major advance is the advent of multislice technology. Multislice CT is a modification of the aforementioned basic configuration whereby multiple detector rows are positioned opposite the x-ray generator. The x-ray generator exposes the subject, and the transmitted x-rays are detected by several rows of detectors on the opposite side of the gantry. Since the first appearance of a dual-slice CT scanner, this technology has burgeoned, and now up to 256-slice scanners are available. The number of detector rows determines the number of images that can be acquired per rotation of the gantry components, meaning that a four-slice scanner can acquire four images per rotation.
In order for horses to be imaged, careful consideration must be given both to the design of the CT table and to the size of the CT gantry opening. The CT table must not only support the weight of the horse but also move with speed and great precision. For a single-slice helical scanner acquiring 1-mm–thick images, the horse must move 1 mm per rotation of the gantry, which generally occurs in about 0.8 seconds. In order to accomplish this, equine CT tables have been custom-designed to work with those manufactured for scanning people. More recently, commercially available equine CT tables have come onto the market (Figure 20-1). An alternative style of CT scanner has recently resurfaced as an option for horses. This type of CT gantry can be equipped with multislice capabilities and is portable. The gantry translates around the outside of the horse with the requisite precision and speed. CT scanners have been built for human applications and as such typically have a circular gantry opening of around 70 cm, which determines the maximum size of the patient or area to be scanned. Currently scanners with 30- to 35-cm openings are commercially available and are appropriate for equine extremity work. There is a movement toward larger gantry sizes that will likely result in machines with >80-cm openings becoming available in the near future, which should increase the clinical utility of equine CT to include larger anatomical regions.
CT images map tissue density in the subject. Each pixel in the two-dimensional CT image actually represents a volume of tissue or volume element (voxel) defined by the pixel size in two dimensions and slice thickness in the third dimension. Typical slice thickness ranges from 1 to 10 mm, and now with multislice scanners, images less than 1 mm thick can be obtained. For each voxel, a value is recorded in Hounsfield units (HU) or CT units. This numerical value represents the density of the tissue within the voxel relative to water, which is arbitrarily set to equal zero. Density values of common tissues are listed in Box 20-1. Lesions can be described as hypodense or hyperdense.
BOX 20-1 Density Values of Different Tissues
|Density (Hounsfield Units [HU])
DDFT, Deep digital flexor tendon.
The viewer, using routinely available image review software, can reversibly manipulate the displayed grayscale value of the tissues. Using a wide “window width” is appropriate for the evaluation of tissues with a wide density range such as bone (approximately 400 to 1500 HU). Viewing with a narrow window produces higher image contrast and is appropriate for the evaluation of tissues with a narrow density range such as soft tissue (approximately 40 to 120 HU). “Level” is to the midpoint of the grey shades that are displayed. Changing the level increases or decreases the brightness of the image. In order to perform a complete and thorough evaluation of CT images, the viewer should actively manipulate both window width and level to view all tissues (Figure 20-2).
Figure 20-2 Transverse computed tomography images all acquired at the same location in the proximal metacarpal region, but processed and displayed differently. Medial is to the left. A has a wide window width and has been processed with an edge-enhancing algorithm to enable accurate assessment of bone. B has the same window, but a smoothing algorithm has been applied for evaluation of the soft tissues. C is a soft tissue algorithm with a narrow window width. D is the same location after the regional administration of contrast medium. This horse has suspensory ligament injury that is moderately contrast enhancing with associated entheous and endosteal new bone at its origin. Note the irregular contour of the palmar cortex of the third metacarpal bone, seen best in A and B, and the periarticular osteophyte bridging the articulation between the second and third metacarpal bones. In C, note the hypodense areas in the suspensory ligament corresponding to areas that show contrast enhancement in D.