Magnetic resonance imaging must be carried out with the patient immobilised using deep sedation or general anaesthetic. Anaesthetic and monitoring equipment must be non-ferrous and MRI-compatible or operated at a sufficient distance from the scanner such that the scanner's fringe magnetic field does not interfere with the equipment nor attract it towards the scanner.


Small animals are usually scanned under general anaesthesia and are intubated and ventilated with isoflurane (which is brain-sparing) vaporised in oxygen. Muscle relaxants are not used but the ventilation (provided by a medical non-ferrous system) is well tolerated and ensures an even plane of anaesthesia. This is essential if the patient is to remain still for long periods and not be disturbed by the knocking noise created the gradient coils during scanning.


Monitoring is performed by means of capnography and blood pressure assessment. If raised intracranial pressure is suspected on clinical grounds, intravenous steroids or mannitol may be given prior to induction of anaesthesia. The morbidity rate is remarkably low given the severe nature of brain disease subsequently diagnosed in some of the patients.

Scanning procedure

For brain scans, the patient is positioned in sternal recumbency with its head in one of the cylindrical RF coils, the smallest coil possible being used for greatest image definition. After an initial procedure during which the coil is " tuned" to reflect its loading by the patient, three single slice " pilot" or " scout" scans are obtained in the transverse, sagittal and dorsal planes. Over these are series of lines are placed on the computer monitor, indicating the location, orientation, number and thickness of the slices in the intended scan.


A standard protocol for brains scans might include:

Transverse T1W scan.

Transverse T2W scan.

Injection of intravenous contrast medium (gadodiamide - OMNISCAN; Nycomed UK Ltd.)

   at the recommended human dose rate of 1 ml/5 kg BW.

Repeat transverse T1W scan to look for abnormal areas of enhancement.

Sagittal and/or dorsal T1W scans - both if a lesion is visible


The T1W scans are always shorter than the T2W scans and with the current system these take 5 minutes and 13 minutes respectively. Total scanning time is usually about 45 minutes. A CSF tap is performed after the scan if the brain has appeared normal or if there is a suggestion of inflammatory disease. CSF removal is not advised in cases of presumed brain swelling or when large masses are present because of the risk of herniation, although lumbar puncture may be safer than cisternal puncture. Recovery from anaesthesia is usually quick and patients are ready to be reunited with their owners within 30 minutes of the study.

Normal appearance of the brain

With T1-weighting, brain parenchyma appears almost homogeneous and of mid-grey signal intensity, slightly paler than the overlying muscle. CSF is dark (hypointense) and bone marrow, subcutaneous and orbital fat is very bright (hyperintense). Cortical bone and air in the frontal sinuses appear as a signal void (black). On T2-weighted scans the brain is seen with a greater degree of contrast, with white matter appearing darker than grey matter and CSF is hyperintense. Surrounding tissues are usually darker and appear grainier than on T1-weighted images. Generally, T1WI show better anatomic detail and are used to visualise enhancement of neoplastic or inflamed areas with contrast medium. T2WI is sometimes a more sensitive detector of minor pathology since increased water content of tissues, which occurs with most disease processes, creates dramatic hyperintense signal. This is why most studies would include a combination of T1- and T2-weighted scans.

Variations in normal anatomy

Brain anatomy does vary slightly between breeds of dog depending on their head conformation. In brachycephalic dogs, the frontal sinus is small or absent and the olfactory lobes lie more ventrally than in longer-nosed breeds. The size of the lateral ventricles of the brain is especially variable and can pose problems in the diagnosis of hydrocephalus. Generally, they are prominent in domed-headed breeds such as bulldogs, Yorkshire terriers and Chihuahuas, possibly due to breeding selection of dogs with appropriate head conformation due to subclinical hydrocephalus. In some patients the septum pellucidum (midline septum between the ventricles) may be lost. The two lateral ventricles may differ in size in some dogs but unless this is associated with a midline shift or other mass effect it is not clinically significant.


Brain tumours

Brain tumours are surprisingly common in dogs and cats. Affected patients are usually middle-aged, but may occasionally be young adults or even puppies. In the dog, meningiomas and gliomas are diagnosed with approximately equal frequency. In the cat meningiomas are by far the most common type of intracranial tumour. The two tumour types can be differentiated with reasonable confidence by their imaging characteristics.


Inflammatory brain disease

Inflammatory brain disease typically produces multiple (" multifocal") lesions of variable size and definition. Modest, ill-defined enhancement may occur on T1W scans but the lesions are usually best seen as ill-defined hyperintense zones on T2W scans. The diagnosis can often be confirmed by CSF analysis.


Congenital brain disease

The most common congenital condition diagnosed is hydrocephalus. May affect the entire ventricular system or only the lateral ventricles. It may be hard to be sure that a moderate degree of ventricular dilation in domed-headed dogs is the cause of the neurological signs because of the "normal" ventricular size variation in these dogs.


Traumatic conditions

Useful modality for assessment of brain trauma provided the patient can withstand the necessary anaesthesia. Cranial fractures, intracranial haematomata, parenchymal haemorrhage and parenchymal oedema can be recognised. MRI can also be used to exclude the possibility of previous trauma as a cause of epilepsy.

Miscellaneous brain conditions

Brain infarcts:

Rarely occur in small animals. Usually seen as wedge-shaped areas of hyperintensity on T2W scans with absent or feeble contrast enhancement on T1W scans. If haemorrhage is present, mixed signal may be seen on both T1- and T2W scans depending on the age of the lesion. Old infarcts may result in brain shrinkage and deviation of the midline towards the lesion.


Acute central vestibular disease (so-called " strokes"): Does not produce recognisable lesions on MRI. MRI is very helpful in differentiating central from peripheral vestibular disease.

Other conditions:

Occasionally, neurological signs may result from expanding masses of surrounding tissue causing compression of underlying brain. Caudally-extending nasal tumours and tumours of the bases cranii and calvarium can easily be recognised even when they are not visibly clinically or with radiography.



Garosi L, McConnell J F, Platt S R, Barone G, Baron J C, de Lahunta A & Schatzberg S J (2006) Clinical and topographic magnetic resonance characteristics of suspected brain infarction in 40 dogs. J Vet Intern Med 20 , 311-321 - PubMed -


van der Merwe L L & Lane E (2001) Diagnosis of cerebellar cortical degeneration in a Scottish terrier using MRI. JSAP 42 , 403-408.


Klopp L S, Hathcock J T & Sorjonen D C (2000) Magnetic resonance imaging features of brain stem abscessation in two cats. Vet Rad Ultra 41 (4), 300-307.


Platt S R et al(1999) Canine intracranial epidermoid cyst. Vet Radiol Ultrasound 40 (5), 454-458.


Le Couteur R A (1999) Current concepts in the diagnosis and treatment of brain tumors in dogs and cats. J Small Anim Pract 40 (9), 411-416.


Forrest L J (1999) The head. Clin Tech Small Anim Pract 14 (3), 170-176.


Thomas W B (1999) Non-neoplastic disorders of the brain. Clin Tech Small Anim Pract 14 (3), 125-147.


Harkin K R et al(1999) Magnetic resonance imaging of the brain of a dog with hereditary polioencephalomyelopathy. JAVMA 214 (9), 1342-1344, 1334.


Snellman M et al(1999) Low-field magnetic resonance imaging of beagle brain with a dedicated receiver coil. Vet Radiol Ultrasound 40 (1), 36-39.


Targett M P et al(1999) Magnetic resonance imaging of a medullary dermoid cyst with secondary hydrocephalus in a dog. Vet Radiol Ultrasound 40 (1), 23-26.


Kuwabara M et al(1998) Magnetic resonance imaging and histopathology of encephalitis in a Pug. J Vet Med Sci 60 (12), 1353-1355.


Taga A et al(1998) The advantage of magnetic resonance imaging in diagnosis of cauda equina syndrome in dogs. J Vet Med Sci 60 (12), 1345-1348.


Saunders J H et al(1998) Probable trigeminal nerve schwannoma in a dog. Vet Radiol Ultrasound 39 (6), 539-542.


Lorenzo V et al(1998) Meningioangiomatosis in a dog - magnetic resonance imaging and neuropathological studies. J Small Anim Pract 39 (10), 486-489.


Ramirez O III et al(1998) A review of imaging techniques for canine cauda equina syndrome. Vet Radiol Ultrasound 39 (4), 283-296.


Kraft S L et al(1997) Retrospective review of 50 canine intracranial tumors evaluated by magnetic resonance imaging. J Vet Intern Med 11 (4), 218-225.


Bayens-Simmond S J et al(1997) Use of magnetic resonance imaging in the diagnosis of central vestibular disease. Can Vet J 38 (1), 38.


Poncelet L.(2011) Epileptic brain damage in dogs and cats: myth or reality? Vet Rec. 169(14):359-60. – PubMed -


Ruth Dennis MA VetMB DVR MRCVS Centre for Small Animal Studies,

The Animal Health Trust, Lanwades Park, Newmarket, Suffolk, CB8 7UU, UK.




Share this with your friends

Click here for the legal bits