Brain structure and epilepsy: the impact of modern imaging.

After the pioneering work of Hans Berger in the 1930s (1), electroencephalography (EEG) opened a new window on epilepsy. This led to its general acceptance and now widespread use. It provided an invaluable new dimension to the ability to locate epileptogenic abnormalities in patients with focal or partial epilepsy. The presence of structural lesions in patients with partial epilepsy had, of course, been long realized but there evolved a widely held concept that the lesion was not nearly as important as the electrographically defined epileptogenic abnormalities. In the early days of neuroradiology, the founding fathers, like Arthur Childe and Donald McCrae, soon realized that asymmetries in skull growth and identification of lesions by pneumoencephalography and arteriography contributed to our study of patients with intractable epilepsy. Nevertheless, imaging continued to labor in the shadow of the EEG until the advent of computed tomography, which ushered in the era of modern imaging in patients with epilepsy (2–8). First came the recognition of disorders or cortical organization such a polymicrogyria and pachygyria (4–7) and the beginning of the explosion of diagnosis by imaging of small inert lesions, some calcified, like the cavernous angiomata (8). It was at that time that we finally realized that electrographic localization, whether from the surface or from intracranial recording, might not provide as reliable an answer as one had anticipated (2, 3). Patients with posterior temporal or temporal occipital lesions and anterior and inferomesial temporal epileptogenic foci had disappointing results after resection of the electrically defined epileptogenic area (9), and it became progressively clearer that the lesion and its immediate surround was much more important in epileptogenesis than we had suspected (8–10). Then came the era of magnetic resonance (MR), which led to important insights into the structural basis of temporal epilepsy, the most common form of the intractable disease, and recognition of a wide range of structural cortical abnormalities, often developmental, leading to seizures which were often difficult or impossible to control. The introduction of MR coincided with a period of renewed and widespread interest in the surgical treatment of epilepsy. Advances in epileptology made it very clear that in as many as 20% of epileptic patients medical control of seizures was not possible with the available pharmacologic armamentarium (2, 3). Advances in surgical technique and, above all, in preoperative electrographic studies and neuropsychological investigation held out the promise of improved results of surgical treatment. There was, in North America, a proliferation of centers hoping to embark on surgical treatment (2, 3). We now see an improved balance between the enthusiasm of the physicians and surgeons and the recognition of the essentialness of searching imaging, EEG, and neuropsychological investigations. Functional imaging is also increasingly used (10). MR imaging is now the standard of reference in the investigation of patients with epilepsy, particularly those with intractable seizures (4). Computed tomography retains a role in recognition of calcifications. It is not a sufficient screening procedure because of the difficulty of recognizing temporal lobe abnormalities and parenchymatous changes. Neurologists and neurosurgeons are now beginning to recognize that referring patients for routine MR imaging is inappropriate. To obtain maximal usefulness, the problem should be discussed in advance with the neuroradiologists and the procedures tailored to the specific clinical problem.