Table 3 Novel imaging modalities for identifying epileptic foci

Magnetoencephalography (MEG) and magnetic source imaging (MSI)

MEG – non-invasive functional imaging recording magnetic flux on the head surface associated with electrical currents in activated sets of neurons. MSI - created when MEG data is superimposed on a MRI [42].

Has been performed experimentally in anaesthetised non-epileptic dog [43].

May be limited by requirement for anaesthesia [44].

Identity microchip may interfere with recording [45].

Requires a magnetically shielded room and other expensive equipment [12].

Positron Emission Tomography (PET)

Functional representational of brain activity (dependent of the radionuclide tracer utilised) e.g. local glucose utilisation (fluorine-18 fluorodeoxyglucose - FDG). Brain regions containing the epileptogenic zone have hypometabolism on inter-ictal FDG-PET [12]. PET and MRI co-registration or integrated PET/MR with simultaneous acquisition is considered superior [8].

FDG-PET may be useful as a diagnostic test for idiopathic epilepsy in the dog [46, 47]

Ictal and inter-ictal single-photon emission computed tomography (ictal/inter-ictal SPECT)

Injection of a radiolabeled tracer during ictus and inter-ictus. Statistical comparison of the blood flow changes. Ideally co-registered to MRI (SISCOM) [48, 49].

Practical difficulties of performing in ictus. Has been performed in inter-ictus and in one study demonstrated subcortical hypoperfusion in epileptic dogs [50]

Diffusion tensor imaging (DTI)

Detects tissue microstructural pathology that influences freedom of water molecular diffusion. Has been used to detect hippocampal and temporal lobe pathology in TLE and DTI tractography has been used in surgical planning [12]. Has demonstrated microstructural alterations in large white matter tracts in idiopathic generalised epilepsy [51]

Experimental studies suggest DTI is feasible in dog [52–54] and structural abnormalities have been identified in a compulsive behaviour disorder [55]. No application for epilepsy yet.

Functional magnetic resonance imaging (fMRI)

Utilises the different magnetic susceptibilities of deoxygenated and oxygenated haemoglobin (blood oxygenation level dependent (BOLD) contrast). Deoxygenated haemoglobin is paramagnetic leading to distortion of magnetic fields and a shorter T2 relaxation time. Areas of increased brain activity have greater metabolic demand and more oxygenated haemoglobin and a prolonged T2 relaxation time. The difference in BOLD at rest and during a specific task (such as language and memory) indicates the areas of the brain activated by the task [12].

Laboratory experimental studies, none relating to epilepsy [56]. Has been used in trained awake dogs to assess cognition [57–59].

fMRI-EEG

EEG is acquired using a specialized system in the MRI machine while acquiring a blood oxygenation level dependent (BOLD) sequence. The EEG is analysed for epileptiform discharges spikes and the corresponding BOLD fMRI change is evaluated [12].

None as yet

Functional connectivity MRI (FcMRI)

Utilizes the principles of fMRI to demarcate brain networks. It evaluates the structural changes distant from the epileptic focus. Main application is in pathophysiology of the epilepsy but has the potential to guide surgery [12].

None as yet

Near infra-red spectroscopy (NIRS)

Probe transmits near infra-red spectrum wavelength rays that passed through the cranium to a depth of approximately 2 cm and is absorbed by haemoglobin in the tissue. Reflected rays are detected by a sensor probe. The strength of reflected rays is inversely related to the concentration of haemoglobin in the brain tissue. The resulting images are co-registered to the MRI to lateralize and localize the signal changes [12].

Pilot studies performed assessing positive emotional states in dogs [60]

Limited to superficial brain structures. May have limited application in dogs with thicker skulls and muscle. However can be performed in awake animals.

Magnetic resonance spectroscopy (MRS)

MRS can be used to measure creatine (Cr), N-acetyl aspartate (NAA), choline (Cho), lactate, myo-inositol and GABA non-invasively in the brain tissue [12]. Reduced NAA/Cho and NAA/Cr was found in the lesional temporal lobe in TLE [61] and in the epileptogenic/irritative zone in frontal lobe epilepsy [62]. These MRS changes were most likely due to cell dysfunction than cell loss [12]

Studies in healthy dogs [63], laboratory canine model of seizures [64] and in some disease states [65].

Arterial spin labelling (ASL)

ASL is a non-invasive MRI technique to assess brain perfusion and therefore image functional areas of the brain. Arterial blood is magnetically labelled using a 180° radio frequency (RF) inversion pulse prior to imaging the region of interest (ROI). The labelled blood flows into the ROI and reduces the MR signal and image intensity at this area. Subtracting this image from the baseline MRI creates the perfusion image which reflects the amount of blood delivered to each voxel [12, 66]. It has been used to show mesial temporal hypometabolism [67] and hippocampal volume loss [68]

None as yet