Diagnostic imaging of injured or ill dogs is frequently required for comprehensive patient evaluation. Veterinary facilities may not be equipped for imaging, or may be limited to plain radiography. Advanced imaging (e.g., MRI, CT) is often ideal, and veterinary facilities do not have these capabilities. This CPG provides guidance for HCPs with extensive training in the use of CT and MRI, when considering advanced imaging requirements, highlighting unique aspects when imaging dogs.
CT is often superior to MRI and used for assessment of margins of osseous or mineralized structures compared to MRI. CT can assess soft tissue changes and differences fairly well by narrowing windows and levels under standard algorithms to see differences of attenuation of the x-rays, but cannot manipulate the soft tissues due to their molecular structure as MRI can in order to enhance or null their differences. Therefore, MRI is often far superior to CT at assessing for subtle changes within soft tissues due to the dramatic contrast enhancement. MRI is most often utilized in veterinary medicine and is the modality of choice when you are trying to assess soft tissue structures not easily accessed by an ultrasound probe or are looking for diseases that may not be appreciated via any other modality. MRI is used primarily for neurologic (brain and spine) imaging and joint imaging concerning cartilage, ligaments, and/or menisci. Keeping those general statements in mind, depending on the type of disease you are assessing for you may be able to appreciate the abnormalities on both modalities, so either study may be adequate for diagnosis. References are provided with specific imaging protocols for MWDs.1-5
The patient must be either heavily sedated or anesthetized while the study is taking place. CT studies of the thorax and abdomen require general anesthesia and intubation of the patient, with closure of the pop-off valve on the anesthetic machine during image acquisition. Depending on how advanced the CT machine is and slice thicknesses needed, this may or may not be a problem for the patient, as the breath hold may have to last for several seconds. Always ensure anesthesia pop-off valves are not left closed, to avoid pneumothorax.
Intravenous iodinated contrast may be used during a CT study in order to further enhance margins of soft tissue structures. If a CT is being conducted to assess an abnormal soft tissue mass or structure, intravenous iodinated contrast should be administered after acquisition of routine images prior to contrast administration for comparison purposes. This contrast administration allows for further characterization of the abnormal soft tissue as only the vascular portions of the structure will enhance.
CT protocols will vary per region you are attempting to image, patient positioning, slice thickness, algorithms, and whether or not contrast will be used. Each of these factors is critical, but the most commonly overlooked factor is patient positioning. Ensure the region of the patient you are imaging is straight and symmetrically positioned on midline of the CT table, as subtle changes in obliquity may make structures appear abnormal when they are not. Use positional aids, sponges, or troughs if needed, and ensure that all metallic or other unnecessary objects are removed. Place the patient either head-first or hindlimb-first into the gantry, depending on which will be closest to the region for imaging. The following are recommended protocols for different body regions based on common problems seen in MWDs.
Patient positioning should be in ventral recumbency, with the hard palate parallel to the CT table. Studies should extend from the tip of the nose to the 2nd to 3rd cervical vertebra. Bone, standard, and bone algorithms with slice thicknesses of 2.5 mm, 1.25 mm, and 0.625 (if available) should be performed, respectively. Sagittal and dorsal reconstructions should be made as needed.
Patient positioning should be in ventral recumbency, with the hard palate parallel to the CT table. Studies should extend from the tip of the nose to the larynx. A bone algorithm with slice thicknesses of 2.5 mm and 0.625 mm (or equivalent) and a standard algorithm with slice thickness of 1.25 mm should be performed. Intravenous contrast should be administered, and the standard algorithm with 1.25 mm thick slices repeated. Dorsal reconstructions are required. Sagittal reconstructions should be made as needed.
Patient positioning should be in ventral recumbency, with the hard palate parallel to the CT table. Studies
should extend from mid-muzzle to the 2nd to 3rd cervical vertebra. Bone, standard, and brain algorithms with slice thicknesses of 2.5 mm, 1.25 mm, and 1.25 mm should be performed, respectively. IV contrast should be administered and brain and standard algorithms repeated. Sagittal and dorsal reconstructions of the standard algorithms are required.
Patient positioning should be in ventral recumbency, with the hard palate parallel to the CT table. Studies should extend from the orbits to the 2nd or 3rd cervical vertebra. Bone and standard algorithms with slice thicknesses of 0.625 - 1.25 mm and 1.25 mm should be performed, respectively. Sagittal and dorsal reconstructions should be made as needed.
Patient should be positioned in dorsal recumbency, with the hind limbs maximally extended caudally (like for a hip-extended VD pelvic view in radiography). Study should extend through necessary vertebral regions based on pain and/or neurolocalization. More specifically for the hind limbs, if UMN signs are present, extend from T8-T9 through sacrum, and if LMN signs present, from T12-T13 through sacrum. CT slices should be acquired perpendicular to vertebral canal (may require gantry rotation). A bone algorithm with 2.5 mm and 1.25 mm slice thicknesses and a standard algorithm with 1.25 mm slice thickness should be performed. For suspect lumbosacral disease, the bone algorithm of 1.25 mm slice thickness should be replaced with 0.625 mm (or equivalent) slice thickness to better visualize the neuroforamina at the lumbosacral junction. Sagittal and dorsal reconstructions of bone and standard algorithms are required.
Anesthesia and breath holds are required. Patient should be positioned in ventral recumbency. Study should extend from thoracic inlet through caudal aspect of liver (ensure extent of all lungs imaged). Bone, standard, and lung algorithms should be performed with slice thicknesses at 5.0 mm, 2.5 mm, and 1.25-2.5 mm, respectively. Sagittal and dorsal reconstructions of lung and standard algorithms are required.
Anesthesia and breath holds are required. Patient should be positioned in dorsal recumbency. Study should extend from caudal margin of cardiac silhouette through pelvic canal (or prostate if male). Bone and standard algorithms should be performed with slice thicknesses at 5.0 mm and 2.5 mm, respectively. Sagittal and dorsal reconstructions of bone and standard algorithms are required.
Patient positioning depends on whether imaging forelimbs or hindlimbs. For forelimbs, the patient is in ventral recumbency. The forelimbs should be extended cranially, resting the forearms and paws on the table with the elbows and shoulders bent at a normal resting position. If the hindlimbs are the focus of the study, the patient is usually placed in dorsal recumbency. The hindlimbs should be placed in maximal caudal extension, keeping both limbs symmetric and including both in the study for comparison purposes (use tape, sponges, or other positional aids). CT slices should be acquired perpendicular to joint spaces, which may require gantry rotation if the joint is the focus of the study. Bone and standard algorithms should be performed along the affected region with slice thicknesses of 1.25 mm. If a joint the focus of the study, conducting an additional bone algorithm sequence with a slice thickness of 0.625 mm is required (if available). Sagittal and dorsal reconstructions of the affected limb only are required.
Magnetic resonance imaging protocols used in veterinary medicine are more simplified compared to human medicine. However, current protocols are adequate in assessing for the majority of diseases of concern.
Use either an MRI-safe anesthetic machine or constant rate IV anesthesia protocols (See CPG 16). Patient monitoring presents challenges in the MR gantry due to increased noise, greater chance of hypothermia, and overall decreased patient accessibility.
It is very important for the Veterinary Corps Officer to be present during image acquisition (if available) to help determine the beginning and end points (range) of the study in each plane, due to anatomic differences between humans and dogs (humans have five lumbar vertebrae compared to seven in dogs, for instance). Beginning and end points for the study should be based on neurolocalization.
Paramagnetic contrast agents are commonly used during MRI. Contrast agent administration is always required when imaging the brain, and may be necessary for other exams dependent on the case. For example, if neoplasia or diskospondylitis of the spine is suspected, then administration of contrast during a spinal exam is warranted. All pre-contrast sequences must be performed prior to contrast bolus administration. The contrast agent most often used in MWDs for MRI is gadolinium-based, and the dose for IV bolus use in the dog is 0.1 mmol/kg (0.2mL/kg). As a quick rule of thumb, 1 mL per 10 pounds body weight is the appropriate dose.
The patient should be positioned in ventral recumbency with the head encased within an effective coil (often head or cardiac types). Studies should extend from the most cranial limit of the orbits/eyes to the level of the 2nd or 3rd cervical vertebra. Slice thicknesses of 3-5 mm should be used; dependent on how many sequences you have time to perform. The following sequences in each respective plane should be performed:
The patient should be positioned in dorsal recumbency, and the coil within the table will likely be used. Study should extend through necessary vertebral regions based on pain and/or neurolocalization. More specifically for the hindlimbs, if UMN signs are present extend from the 8th or 9th thoracic vertebra through the sacrum, and if LMN signs are present, from the 12th or 13th thoracic vertebra through sacrum. Slice thicknesses of 2-4 mm should be used; dependent on how many sequences you have time to perform. The following sequences within each respective plane should be performed:
The patient should be placed in lateral recumbency, with the affected limb up, with the stifle placed in neutral to moderate extension. Study should at least extend from distal femoral diaphysis to the proximal tibial diaphysis, distal to the tibial crest. A wrist coil is preferable, however if the joint/region to be imaged is too large, then cardiac or other similar coils may be used. Slice thicknesses of 2-3 mm should be used; dependent on how much time you have to complete the study. The following sequences within each respective plane should be performed: