Moving away from the use of general anaesthesia for diagnostic imaging in equine practice is an ongoing challenge. The risk of related injury, myopathy and neuropathy cannot be an overlooked when evaluating the use of new technology. Although marginal advancements in pre-operative and procedural care have been made, the associated risk of anaesthesia in horses is still disproportionally high in comparison to other companion animals (Dugdale, 2016). There are also staff safety concerns to consider in relation to equine computed tomography (CT). The large volume of radiation emitted during scanning increases occupational exposure and subsequent risk to staff, preventing staff from manually restraining patient during the procedure. Many methods have been adapted recently, with these key factors considered, to ensure the continuation of high image quality and diagnostic accuracy. Within this article, some of the different systems that have been created to overcome the complexities faced by those working in the equine veterinary industry are reviewed.
Computed tomography hardware
CT imaging is built up of three main components; the patient couch, x-ray tube and the detectors, which rotate inside the gantry around the patient, acquiring and collecting attenuation data. The computer, which processes attenuation data, uses pre-defined reconstruction algorithms to create images. The user also has access to the operating console which enables software access, gantry control, exposure factor control and post-processing viewing, enabling the manipulation of data to ensure attainment of imaging is suitable for purpose. This also allows for the construction of 3D imaging.
Typically, veterinary practices will also utilise a picture archiving and communication system (PACS), enabling the storage of all images gained from separate modalities in a central location. The common format of image archiving is digital imaging and communication (DICOM), which ensures images are readily available, transferrable and may be viewed and manipulated via a DICOM viewer installed on any computer. Although, it is worth noting that the screen resolution may limit image quality (Seeram, 2018).
Conventional CT makes use of a fan shaped beam of x-rays collecting a series of axial slices in a continuous spiral (helical) motion about the subject. This is achieved by the patient either moving through the aperture of the gantry or the gantry of the system moving along the axis of the patient (Garvey, 2002). Cone beam CT (CBCT) uses a cone-shaped beam of x-rays and a flat panel detector mounted opposite, using as minimal as one rotation about the subject (Venkatesh, 2017). Multi-slice helical CT (MSCT) and fan beam geometry is typically used in CT scanners available to the veterinary market, because of their fast image acquisition and lower doses of radiation and chemical restraint (Power et al, 2016). MSCT uses slip ring technology, where electrical energy is passed to a rotating ring using brushes so no static wires exist between the two enabling a continuous rotation. This, coupled with moving the patient through the gantry aperture, enables the helical data set to be acquired. The slice geometry of MSCT does not create a single planar section of data so a step during algorithm reconstruction, called interpolation, must be used. Interpolation is a mathematical technique for estimating the value of a function from the known values on either side of it (Seeram, 2018).
Advantages of computed tomography
Radiographic examinations are typically the first imaging modality of choice for suspected bone-related pathology in the horse. Superimposition and obliquity of complex structures restricts evaluation of all potentially affected areas, such as skull fractures. The use of a cross sectional imaging modality like CT enables further information which can then be considered for surgical planning and prognosticating a case (Crijns et al, 2019).
CT scanning uses x-ray beam attenuation as the fundamental for the image formation, the image created uses a grey scale that is assigned to the varied tissue density, dependent on the CT number (Hounsfield unit) of the tissue. Moreover, CT produces user friendly images, the interpretation of which is similar to radiography where tissues of low density, such as an air-filled lung, will appear dark and are described as hypoattenuating, whereas tissues such as enamel or bone will appear bright, and are described as hyperattenuating (Labruyère and Schwarz, 2013). With MSCT, small (<1mm) and over-lapping slices can be obtained, which contributes to the high spatial resolution and ability of CT to be used in multiplanar reconstruction and 3D surface reconstructions (Puchalski, 2012).
Limitations of computed tomography
Although CT has many benefits in practice, including fast image acquisition, it is not a flawless imaging modality. CT imaging is affected by at least five physical properties:
- Spatial resolution
- Contrast resolution
- Temporal resolution
- Noise
- Artifacts.
Exposure factors are directly linked to the quality of the images. Increasing mA (beam quantity) and increasing kV (beam quality) reduces noise and helps to improve image quality but will increase effective dose of radiation (milliesievert) to the patient. Operators are committed to working within the remit of the ‘as low as is reasonably achievable’ (ALARA) principle, whereby patient dose is kept to a minimum but the image quality is not compromised (Seeram, 2018).
CT has limited efficacy in the identification of soft tissue lesions because of the way in which images are created, despite algorithms, window level and width settings for soft tissues. When two tissues of similar density are in apposition, it can be hard to differentiate between them. Practitioners can utilise contrast enhancement to help alleviate this problem (Puchalski, 2012).
Contrast media
Iodine-based contrast media is typically used because of its high atomic number, which enhances attenuation. Intravenous (IV) administration of contrast highlights alterations to blood flow, which can occur within injured or inflamed soft tissue, and this broadens the scope of CT capabilities in relation to soft tissue lesions (Nelson et al, 2017). Comparison between intra-arterial (IA) and IV contrast injection for delineating structures in the head identified similar enhancement of structures, when compared to pre-contrast images. Although both methods promote definition, IA has the potential to evaluate lesions of increased blood flow, whereas IV provides a more homogenous and symmetrical enhancement (Crijns et al, 2016).
IA contrast has commonly been used to identify activity of deep digital flexor tendon (DDFT) lesions. Typically images are acquired pre-contrast, then the medial artery of both forelimbs are catheterised and iodinated contrast is administered via pressure injectors set to 1.5ml/s and started 10 seconds before acquisition, allowing perfusion (Wilson et al, 2021). Contrast enhanced CT (CECT) has previously been used to identify lesions of the DDFT and characterise their activity, delineating neovascularisation and pinpointing the areas of interest which may then be reviewed via magnetic resonance imaging (MRI) and histopathology to confirm the margins of the lesion (Puchalski et al, 2009). Intrasynovial use of contrast, arthrography, tenography or bursography, can be used to outline the surface of structures that comprise the region. Hontoir et al (2014) reported that CT arthrography was better than high field MRI for identification of cartilage lesions in the equine fetlock joint, as seen in Figure 1.
Although limitations exist for the assessment of articular cartilage with both MRI and CT, for example the low spatial resolution and long acquisition time of MRI and the low contrast resolution of CT, pre-operative CECT arthrography can enhance procedural planning and has been used to demarcate osteochondroma of the distal calcaneus (Nelson et al, 2018). De Souza et al (2020) used intra-articular contrast within the tarsocrural joint to delineate four separate mineral structures surrounded by a soft tissue capsule; suggestive of a singular mass, as seen in Figure 2.
CT tenography of the digital flexor tendon sheath of the hindlimb has also been described as a superior diagnostic test, where ultra-sound and contrast radiography have failed to give an accurate diagnosis before tenoscopy. Agass et al (2018) reported that CT gave excellent anatomical detail of the intrathecal structures, including the borders to the digital flexor tendons, 100% of the time and the manica flexoria and plantar annular ligament 95% of the time (Agass et al, 2018). Dacryocystography can also be performed to highlight the margins of the nasolacrimal duct in patients with suspected pathology of the area. The contrast can help to ascertain patency when drainage issues are suspected (Figure 3) (Nykamp et al, 2004).
Adverse events associated with use of contrast media, including hyperthermia, urticaria and skin oedema have been noted in previous equine studies. Research reviewing associated reactions in general anaesthesia-induced myelographic procedures found that 34% (n=95) of 278 patients had an adverse reaction, and of those 95 patients, 5% were euthanised as a result of the adverse reaction and 3% were euthanised at the owner's discretion. The adverse reactions were significantly associated with larger volumes of contrast material and longer time spent under general anaesthesia, when compared to patients that did not show any signs of an adverse reaction (Pollard and Pulchalski, 2011; Mullen et al, 2015). Another study by Gough et al (2020) found a similar incidence of post-myelographic contrast reactions with 36% of the 51 cases reviewed displaying an adverse reaction. No horses were found to display any clinical signs of seizure like activity or require euthanasia, although one horse did suffer a catastrophic fracture during the anaesthetic recovery period. The authors concluded that the nature of the procedure was in line with previously described risks of any general anaesthetic procedure and that the superior diagnostic and prognostic information that CT provides outweighed the risks (Figure 4) (Gough et al, 2020).
Clinical applications: standing and general anaesthesia
CT systems for equine patients are becoming more widely available throughout the UK at specialist centres, for both standing sedated and/or general anaesthesia patients. Depending on the practices requirements, there are different systems commercially available with-in the UK, many of which utilise one gantry with an interchangeable patient bed and standing head scanner (Figures 5 and 6). Most systems are developed with the practical applications of the set up in mind, and are individualised to the centre's requirements, so variations may be seen across different venues.
Standing sedated CT in equine patients can be used for imaging the head, most commonly to look for pathology within the paranasal sinuses, nasal cavities, larynx, pharynx and guttural pouches, dental arcades, orbit, temporomandibular joint and some of the cranial nerves (as accurate identification of all cranial nerves from their surrounding tissues is not possible with current CT protocols.) It is also an invaluable tool for evaluation of head trauma and skull fractures (Dakin et al, 2014; Dixon et al, 2017). One of the most common uses of standing head CT is assessment of the dental structures. CT is considered the gold standard imaging modality for teeth and the surrounding structures. Assessment of the enamel, infundibulum, pulp horns, periodontal space, alveolar bone, lamina dura and root changes are all possible with CT (Bühler et al, 2014; Manso-Díaz et al, 2021). A comparative study found that CT had higher sensitivity (100%) and specificity (96.7%) in identification of dental and sinonasal disease compared to radiography at 72.5% and 89.5% respectively. For osseous cases three facial fractures were compared, and in all cases CT was able to identify extra bone fragments that were not visible radiographically. CT was also able to identify fractures in complex regions such as the zygomatic and temporal bones (Manso-Díaz et al, 2015). Imaging the cranial cervical vertebrae is possible standing, although by using this method there are quality limitations because of the potential for reduced patient cooperation, movement artifacts and patient size issues.
Two main techniques have been described, which include a sliding gantry system that passes over a stationary head or a stationary gantry where the patient's head is passed through the aperture of the gantry. The latter requires the patient's body to be suspended on an air platform hovercraft system designed to make the patient ‘weightless’, it can then be driven by the couch mechanics. Typical scan times are approximately 30 seconds, as this enables acquisition of attenuation data from C1 to the incisive region of the skull (Figures 5 and 6) (Dakin et al, 2014).
Standing limb CT, although in early development, has the potential to revolutionise orthopaedics and lameness diagnoses in the equine patient. Many authors, from different centres, have reported various means of acquiring standing CT images of the distal limbs. One such system available uses CBCT technology and operates robotic arms mounted with an x-ray tube and flat panel detectors that are controlled to move around the area of interest systematically to acquire the images (Riggs, 2019).
A system available through Asto CT™ enables patients to walk over the gantry before raising up around the limb. The attenuation data is gathered for both fully weight bearing limbs and is generally capable of attaining images from the hoof to the proximal carpus or tarsus with a 75 cm gantry bore and imageable field of view (Figures 7 and 8), often within 30 seconds. Another benefit can be observed during standing surgical procedures, as patients can stand on the platform, allowing for CT imaging to occur at intervals of the procedure; checking placement and progress. Very low tube currents are used, so a minimal dose is produced, enabling operators to stay safely with the patient. The gantry has been designed so that it is sealed and can withstand a 1000 kg strikeforce should an incident occur. The operator control panel has also been created with veterinary practice in mind and easy to use software.
Another system commercially available through Vet-DIcon in Germany is the Qalibra™. Joint partnership between Canon Medical, using their Aquillion LB gantry system with an opening of 90cm and imageable field of view of 70cm, has been linked to a patented sliding and height-adjustable platform controlled via the CT control panel. The gantry is lowered to floor level and the patient places a fore- or hindlimb through the aperture of the gantry; completing the scan by the gantry moving away from the patient during attenuation data collection (Figure 9).
At the end of 2020, Hallmarq© launched a new CBCT standing system capable of imaging the distal limb (Figure 10, 11and12). The system is installed at sub floor level or within a low platform frame, prioritising horse safety by creating a walk in/walk out set up. The detector plate and generator revolve around the limb on a novel dual concentric ring design. The plate remains close to the limb, improving image quality, with an acquisition time of 60 seconds. Additionally, Hallmarq is one of the few companies to incorporate motion correction technology into their software, to better ensure clear high-quality images of the standing patient. However, this technology is still in the early stages of development and continues to be modified.
General anaesthesia CT is less widely available in the UK compared to standing CT. However, it is invaluable for imaging of the entire cervical vertebrae and cranial thoracic vertebrae. Cervical vertebral column pathology has been identified as the leading non-infectious cause of ataxia in equine patients (Lindgren et al, 2020). CT scanning of the cervical vertebrae has become an additional next-step diagnostic tool after traditional latero-lateral and laterodorsallateroventral radiographic projections have failed to give an accurate diagnosis and a lack of detail (Withers, 2010). Radiographs of the articular process joints (APJ) using both projections can provide information as to asymmetric APJ enlargement. Axial enlargement of the APJs can cause compression of the spinal cord, whereas ventral enlargement expands into the intervertebral foramen in which the spinal nerves pass (Biggi et al, 2018). The tomographic nature of CT has the advantage of removing superimposition and allowing multi-planar reconstructions, improving the possibility accurate diagnosis because of its higher sensitivity and specificity (Gough et al, 2020). CT is commonly used in combination with myelography as a diagnostic tool, with the potential to become a gold standard diagnostic modality of the future with further refinement (Figures 13 and 14) (Dixon, 2018).
CT myelography is indicated when compression of the spinal cord is suspected. Although identification of vertebral column narrowing is possible on radiographs, it is not possible to ascertain the degree of spinal cord compression without the aid of positive contrast delineating the margins (Biggi et al, 2018). Most neurological defects attributed to the cervical region are compressive. Cervical vertebral compressive myelopathy (CVCM) is the most common, caused by vertebral arch malformation or pathology of the APJs. It has been described in traditional radiographic myelography, where patients are positioned in lateral recumbency and lateral-lateral projections are acquired in neutral, flexed and the extended position, that it may be complex to determine the significance of dorsal ventral vs dorsal-lateral compression (Hepburn, 2018). One study suggested that a 50% reduction in the contrast column, both dorsally or ventrally, was not a reliable indicator for compressive lesions (Hahn et al, 2008). CT myelography enables axial slices of the subarachnoid space to be acquired, and the circumferential visualisation of the dural and spinal margins allows differentiation between the cause of CVCM. One study with a sample population of 51 horses identified spinal cord compression to be present in 61% and 49% of those being classified as lateral or dorso-lateral compression, respectively (Gough et al, 2020). However, one prominent limitation of a CT myelography is access to anatomy. Owing to the size of the patient and gantry, only images of the cervical spine in a neutral position can be obtained, so currently supplementary radiographs to collect data on the flexed and extended images are still needed (Gough et al, 2020).
CT imaging of the thoracic, abdominal and pelvic regions in most average size horses is not feasible, although with larger bariatric CT scanners coming to the market, this has been achieved on smaller patients. Imaging of these regions in smaller breeds and foals is readily available, with evidence to demonstrate the validity of its use in these instances. Case selection is commonly associated with traditional modalities failing to identify pathology to correspond with the clinical signs. Because of this, many centres rarely use CT for in this context, so CT of these regions in relation to poor performance is yet to be used (Schliewert et al, 2015; Powell, 2018). The normal appearance of thoracic CT in foals has been described post-contrast enhancement, then compared to postmortem analysis, to further identify thoracic structures on CECT. This research serves as an anatomical reference and method for clinicians to allow interpretation of future cases of suspected thoracic disease in foals (Arencibia et al, 2020). Abdominal CT of foals has also been described for the identification of a portosystemic shunt using CECT to identify the hepatic vessels. A small vessel was seen to arise from the portal vein in a caudal direction entering the left side of the caudal vena cava (Willems et al, 2019). Diagnosis of pelvic fractures in horses usually relies on a combination of clinical examination, scintigraphy, ultrasound and radiography, but CT has been utilised in foals with suspected pelvic fractures successfully. The ability to accurately assess fracture configuration enables better prognostication for long term athletic function for these patients (Ducharme and Nixon, 2019).
General anaesthetic CT limb imaging is typically carried out using a modified patient bed. Attachments to the human coach or system available via Vet-DIcon have been modified to withstand the weight of the horse. The gantry moves over the patient, obtaining data within approximately 10 minutes. Both systems image from hoof structures all the way to the carpus or tarsus, with some centres reporting imaging of the stifle and pelvic regions (Figure 15) (Bergman et al, 2007).
With the development of larger bore gantries, proximal limb imaging has become more attainable; with the added benefit of reducing patient risk, as less unnatural positions may now be utilised. As the complex stifle joint comprises multiple compartments, including the menisci and supporting ligaments, a multimodal imaging approach is commonly needed to evaluate the joint, but even with this approach there are limitations to lesion identification (Adrian et al, 2017).
Although open low field MRI systems are available and considered a ‘gold standard’ technique to allow further imaging of the stifle joints they require prolonged anaesthetic times, requiring two anaesthetics if contralateral limb or arthroscopy is required (Waselau et al, 2020). Stifle CT with contrast enhancement is much faster and can allow imaging of osseous and soft tissue structures of the joint as well as arthroscopic treatment if required in the same general anaesthesia; reducing patient risk (Figure 16 and 17) (Vanden-berghe, 2020).
Owing to the sensitivity of CT, previous case studies have been able to identify palmar metacarpal stress fractures where other tools, such as radiography and ultrasound, have failed to give a diagnosis (Beccati et al, 2019).
CT imaging has also been compared to the widely available low field MRI system. A study of hoof scans of 31 limbs from 22 horses using both modalities, compared lesion detail and concluded that MRI was superior when imaging DDFT lesions within the distal to the proximal portion of the navicular bone. However, CT was able to identify lesions more proximately, that were not visible on the MRI (Vallance et al, 2011). This was attributed to a smaller field of view and signal drop out of the proximal portion of the MRI protocol. Compared to MRI's ability to identify physiological properties of the body, including fluid accumulation within bone or ‘bone oedema’ CT is at a slight disadvantage (Vallance et al, 2011). However, new techniques from human medicine using non-calcium bone reconstructions from dual-energy CT have been adapted by Vet-DIcon, which have an adequate sensitivity and specificity for the diagnosis of bone marrow oedema, highlighting scope for future improvements (Diekhoff et al, 2019). CT can also be used with positron emission tomography (PET), as a way of identifying the metabolic activity of lesions within bone and soft tissue. This requires the patient to be injected with a fluorine-based radioactive tracer. Images are then acquired separately with a PET and CT scanner and retrospectively fused together. Although this is still in the early stages of veterinary clinical use, an exploratory study did find that PET scanning was in agreement with CECT and T2-Weighted and Short Tau Inversion Recovery, also known as fat-suppressed sequence, scans on MRI for the identification of DDFT lesions in the hoof (Wilson et al, 2021).
Conclusion
Over the last few years there have been many advances in the use of CT in equine patients. These have progressed from all imaging being done under general anaesthetic with adapted patient beds, to standing head CT being acquired using a patient hovercraft system. More recent changes to bariatric scanners have enabled larger portions of the equine patient to be imaged as well as the use of CBCT systems allowing for standing limb imaging. Many companies have strived to develop a multi-functional CT scanning model for equine practice, and many have showed great promise for the suitability of imaging from the hoof to carpus or tarsus, as well as the head and cranial neck. Although, with few installations currently, the overarching success of these models are yet to reach their full potential.
KEY POINTS
- Computed tomography technologies have developed rapidly in recent years, with an uptake in interest of the modality from the equine industry many computed tomography scanners have been designed with veterinary practice in mind and continue to be developed still.
- Computed tomography scanning of the equine patient produces many constraints regarding getting the patient to the scanner as well as keeping the patient within the gantry during image acquisition.
- Issues regarding the use of equine computed tomography have to be overcome including staff safety with radiation, patient safety whether under general anaesthesia or standing sedated.