20.3.1 Seizure Semiology

Seizure semiology is the clinical manifestations of habitual seizures of the patient and it has an important lateralizing and localizing value. For example, temporal lobe and temporal plus epilepsy are typically seen with auras, vegetative or visceral symptoms such as behavioral arrest, staring spells, vague epigastric sensation, nausea, auditory or visual hallucinations, dream states, and/or language dysfunction. On the other hand, seizure semiology in extratemporal epilepsy is less stereotypical and much more variable. It changes widely based on the location of seizure onset and its propagation pathways. Typical frontal lobe seizures are seen with motor automatisms, dystonic posturing, head and eye deviation, tonic, clonic, or atonic activity, drop attacks, olfactory manifestations, vocalization and complex behaviors. Parietal lobe seizures are frequently associated with somatosensory, abdominal symptoms (nausea or choking), and gustatory sensations and they spread to frontal or temporal regions easily. Occipital lobe seizure symptoms are mostly present as visual phenomena, hallucinations, flashes, scotomas, hemianopia, contraversive and ipsiversive eye and head movements. All these semiological characteristics of the seizures help to locate the epileptogenic zone and accurate description of clinical semiology can be most reliably obtained through long-term video EEG monitoring, ^{8 , 10 , 11 , 12}

20.3.2 Electroencephalography

Electroencephalography (EEG) remains the cornerstone of preoperative workup in determining the epileptogenic zone. Electrophysiological data may be obtained with non-invasive techniques such as routine EEG, long-term scalp video EEG monitoring (LTM), and magnetoencephalography (MEG). However, routine EEG may be normal in about 30–50% of the patients with epilepsy. ^{8 , 9 , 10 , 11 , 12 , 13} Therefore, although abnormal interictal EEG findings are helpful, a normal EEG does not rule out presence or location of epilepsy. The value of ictal EEG is much higher and documenting the abnormal electroencephalographic activity during the onset of habitual seizures constitutes the single most significant information. LTM plays the most critical role in obtaining both ictal and interictal EEG recordings and also documenting the clinical characteristics of habitual seizures. Correlation of clinical semiology and abnormal EEG findings are essential to lateralize and localize the epileptogenic zone. LTM with scalp EEG is a routine part of the preoperative assessment of surgical candidates in Phase I and might be sufficient to select surgical candidates in many cases. However, if the non-invasive electrophysiological techniques fail to define the epileptogenic zone or the results from diagnostic work up are discordant, then invasive monitoring techniques are used to locate the epileptogenic zone. Invasive LTM is performed by using surgically placed intracranial electrodes such as subdural strip, grid and intraparenchymal depth electrodes. These electrodes can also be used to map cortical functions by stimulating cortical and subcortical areas. ^{8 , 9 , 10}

Invasive monitoring is indicated if:

• 
  

  surface EEG data is inconclusive to lateralize and/or localize the epileptogenic zone,

  
  
• 
  

  clinical, electrophysiological, and radiological data is noncongruent,

  
  
• 
  

  multiple epileptogenic areas or structural abnormalities are present,

  
  
• 
  

  there is a single structural lesion with ill-defined borders,

  
  
• 
  

  MRI is negative despite electrographically well-documented epileptogenic activity in a certain cortical region,

  
  
• 
  

  epileptogenic zone extends in or abuts the eloquent cortex,

  
  
• 
  

  suspicious epileptogenic zone is in a deep-seated location.

The goal of invasive monitoring is precise determination of the epileptogenic zone, seizure propagation pathways and mapping of the eloquent cortex. It provides a reliable electrophysiological recording directly from cortical surface or deep structures. However, the goal can be achieved only after determining the appropriate coverage area to place intracranial electrodes. Coverage area is defined based on a hypothesis about possible location or network of the epileptogenic zone. Non-invasive electrophysiological data, imaging studies, Positron emission tomography (PET), Single-photon emission computed tomography (SPECT), and clinical semiology are used to develop a reasonable hypothesis regarding the possible location of the epileptogenic zone and propagation paths before proceeding to invasive LTM.

The most commonly used intracranial electrodes are subdural strip, grid and depth electrodes. They have different indications, advantages and disadvantages. Strip electrodes are single row of electrode arrays imbedded in thin silastic sheets and can be placed easily through a simple burr hole. They are ideal to lateralize the seizures. Grid electrodes are rectangular arrays with several parallel rows of contacts. They cover a larger area and are ideal to localize and map the borders of epileptogenic zone and adjacent eloquent cortex. However, relatively large craniotomies are needed for subdural grid placement. Depth electrodes are multi-contact electrode arrays embedded in a very thin, tubular silastic material. They provide excellent recordings from deep structures such as the amygdala, hippocampus, cingulum, and orbitofrontal cortex. They are placed through small drill holes using stereotactic technique. The most common application of depth electrode is hippocampal placement and stereo-electroencephalography (SEEG). ^{9 , 10}

Placement of subdural electrodes has a higher risk for CSF leak, infection, hemorrhage, cortical injury and cerebral edema. Although these risks are much less with depth electrodes, there are some disadvantages of depth electrodes. The main disadvantages are it provides minimal cortical coverage, limited functional stimulation capability and limited cortical sampling (tunnel vision). Selection of the electrodes should be determined by coverage strategy. If the purpose is covering an electrographically well-defined epileptogenic lesion and adjacent eloquent cortex, then subdural grid electrode might be the best option for cortical EEG recording as well as stimulation and mapping of the cortex. If the epileptogenic zone is not well defined and multiple potential epileptogenic areas are present or the epileptogenic zone is deep seated, then stereotactically placed depth electrodes are more appropriate to explore the epileptogenic network. ^{9 , 10}