• epilepsy
• neuroimaging

The value of magnetic resonance imaging (MRI) of the brain in people developing epilepsy has been amply demonstrated.¹ This is particularly the case for those with medically refractory partial seizures, for whom appropriate surgical treatment may result in complete resolution of seizures in 70%. MRI scanning has become much more available over the last decade in many community hospitals, to the benefit of the populations they serve. There are some concerns, however, about quality control, and particularly because of the shortage of trained neuroradiologists. This important issue is addressed in the paper by Von Oertzen et al (this issue pp 643–647).²

It was found that MRI scans carried out and reported on in German community hospitals had a low rate of detection of focal lesions such as hippocampal sclerosis, tumours, cortical malformations, and vascular anomalies. There was a modest increase in detection rate by having these scans reviewed by expert neuroradiologists at the Bonn Epilepsy Centre. There was a greater increase in detection by obtaining epilepsy dedicated MRI in the centre, and focal lesions were identified in 85% of patients with unremarkable standard MRI.

Does this mean that all patients with refractory epilepsy need to be referred to specialist centres for their MRI scans? At present this would be an appropriate part of the evaluation for possible surgical treatment of the epilepsy, but it is also important to improve the diagnostic accuracy of MRI scans obtained at smaller units. There is some impact from improved scanning hardware, but the greater advance is possible from the use of MRI acquisition protocols for patients with epilepsy that may be used by radiographers without extensive specialist training. Protocols have been formulated in guidelines by the International League against Epilepsy^3,4 that have also considered the complementary role of functional imaging with MRI.⁵ A typical basic epilepsy MRI protocol would comprise a T1 weighted volume acquisition which may be reformatted in any orientation and used for volumetric measurements: proton density, T2 weighted, and fluid attenuated inversion recovery (FLAIR) acquisitions obtained with contiguous slices that cover the entire brain with as thin slices as possible and orientated perpendicular to the hippocampus. The oblique coronal orientation minimises partial volume effects that otherwise commonly obscure hippocampal sclerosis and small lesions in the temporal lobe, and also displays the rest of the brain in a clear and logical fashion. Standardisation of acquisition protocols also facilitates subsequent review of images at a specialist centre and may avert the need for repeat scans.

There is a clear need for an increased number of trained neuroradiologists to direct image acquisition and to report on MRI scans. In the meantime, current resources could be used more efficiently by the increased use of standardised acquisition protocols and effective telereporting in the managed clinical networks of community hospitals and regional specialist centres.