Blume's Atlas of Pediatric and Adult Electroencephalography, 1st Edition

Chapter 6

EEG in the Intensive Care Unit

The intensive care unit (ICU) is a new frontier for EEG, posing opportunities and challenges. The application of EEG in ICUs has been facilitated by the advent of digital EEG with large storage capacity, high-frequency sampling, and networking capability. From clinical neurological and clinical neurophysiological perspectives, EEG is long overdue in this application, as it allows real time, noninvasive, safe, inexpensive, and sensitive assessment of cerebral cortical function at the bedside. Cortical function cannot be clinically assessed when patients are obtunded from disease or drugs or are paralyzed from disease (e.g., neuropathies) or neuromuscular blocking agents, yet the EEG provides a window to cortical activity and integrity. For the neurological consultant, EEG extends the assessment of brain function above the brainstem in the unresponsive patient and provides an assessment of cortical–thalamic function. Of course, this electrophysiological information must be considered in the light of the clinical picture, the confounding effects of drugs on the EEG, and potentially misleading artifacts peculiar to the ICU. The type of EEG (continuous or brief/sporadic, quantitative or raw, full array or “subhairline”) must be considered carefully along with the question to be addressed. Applying EEG in a thoughtful manner makes it a useful tool that complements the clinical assessment, imaging and other laboratory tests.

EEG can be used to detect and monitor the treatment of seizures, ischemia, cerebral edema, and even developing mass effect of intracranial hematomas and other lesions before severe damage is done. There is good evidence that seizures other than absence can contribute to brain damage (Nairismagi et al., 2004; Young and Jordan, 1998; Young et al., 1996). Many of these seizures are nonconvulsive and cannot be reliably detected clinically. EEG shows changes due to ischemia at levels of regional/total cerebral blood flow that are well above the threshold for infarction (Jordan, 1995). The effects of treatment can also be monitored, making a perfect “closed system.” We have used EEG successfully to monitor depth of sedation in paralyzed patients (Savard et al., 2009) and to grade the severity of encephalopathy, noting worsening or improvement for broad classification, e.g., differentiating seizures from metabolic encephalopathies (Young et al., 1992; Young, 2000).

EEG in the ICU requires familiarity with artifacts, some of which are peculiar to the ICU. These are reviewed in Chapter 9, under “Recording in an Intensive Care Environment.”

The presence of absence of variability and reactivity are of great prognostic importance, especially for traumatic and anoxic–ischemic encephalopathies (Al Thenayan et al., 2009). Methods of testing for reactivity are also discussed in Chapter 9, “Recording in an Intensive Care Environment.”

Some of the challenges of ICU EEG include:

  1. Asking the right question. What does the clinician need to know? For example, sedation monitoring can be done with an EEG of four or fewer channels. For seizures, long-term monitoring (LTM) for at least 48 hours is optimal (Claassen et al., 2004). Is there a differential diagnosis to be made? The EEG can help make broad differential diagnoses, e.g., in differentiating seizures from metabolic encephalopathies. Is there a prognostic issue for which trending or serial EEGs may be suitable? There is some promise for the role of EEG in assessing comatose victims of cardiac arrest. Sporadic EEGs can be of value, but there is more prognostic promise in “trending” the automated or raw EEG over 48 h or more to determine whether there is variability (invariant EEGs carry a worse prognosis), improvement, or worsening (Hebb et al., 2007).
  2. Which patients to monitor? Some groups—e.g., those with recent convulsive seizures and those with acute structural brain lesions—are at high risk for seizures (Young et al., 2005).
  3. Who reads the EEG? For long term monitoring (LTM), this is a practical issue. Some technologists and EEGers have taught the bedside nurses; other neurologists read the EEG sporadically during the day. Automated programs can help in seizure detection; for example,

P.588


changes in power spectral analysis or in alpha/delta ratio can be used to detect seizures (Vespa et al., 1999). These must be confirmed by referring to the “raw” EEG, but the data reduction is very helpful, as LTM records are extremely time-consuming to read.

  1. Developing a “system.” LTM usually requires regular checks of electrodes, networking the EEG, which can be read remotely (as EEGers and EEG technologists cannot usually stay at the bedside), a method of communication of the results of the EEG to the responsible physicians, and follow-up for further monitoring and feedback until the problem is resolved.
  2. Artifact recognition. Many artifacts are peculiar to the ICU; these occur without the watchful eye of the EEG technologist, at least for LTM cases (Young and Campbell, 1999). They are discussed in Chapters 2 and 9.
  3. Waveform recognition: there are still difficulties in defining electrographic seizures in the ICU and in differentiating triphasic waves from spike–waves.

We have developed a reliable classification system for grading the severity of encephalopathies (Table 6-1) (Young et al., 1997). Representative examples will be shown. In addition, some features of focal disease, including asymmetries will be demonstrated.

Table 6-1 EEG Classification for Encephalopathies

Category

Subcategory

Delta/theta >30% of record (not alpha–theta pattern coma)

A. Reactivity

 

B. No reactivity

Triphasic waves

Burst suppression

A. With epileptiform activity

 

B. Without epileptiform activity

Alpha/theta/spindle coma (unreactive)

Epileptiform activity (not in burst-suppression pattern)

A. Generalized

 

B. Focal or multifocal

Suppression

A. <20 but >10 µV

 

B. ≤10 µV

This table lists the principal categories of EEG abnormalities found in encephalopathic ICU patients. The categories are presented roughly in ascending order of severity.

 

Fig. 6-1. No caption available.

 

Fig. 6-2. Mild dysrhythmia. Patient's age, 65 years. Stuporous. There is diffuse, variable, symmetrical theta of low-medium voltage. The patient had hypercalcemia secondary to hyperparathyroidism. The EEG and the patient's level of consciousness improved with correction of the hypercalcemia. Calibration signal 1 s, 70 µV.

 

Fig. 6-3. Moderate dysrhythmia. Patient's age, 76 years. Comatose. There is diffuse arrhythmic delta and theta. Note the faster frequencies are relatively less developed in the left hemisphere. There is good variability with attenuation of delta and an increase in faster frequencies in the last 1.5 seconds. The patient was ill and toxic with Group B streptococcal endocarditis and, in addition, suffered an embolic infarction in the left middle cerebral artery territory. The patient ultimately survived and made a favourable recovery after rehabilitation. Calibration signal 1 s, 70 µV.

 

Fig. 6-4. Moderate dysrhythmia. Patient's age, 71 years. Comatose. Intermittent rhythmic delta alternating with runs of theta and low voltage delta activity, slightly higher on the right. The patient was recovering from coronary artery bypass graft. Calibration signal 1 s, 50 µV.

 

Fig. 6-5. Moderate to severe dysrhythmia. Patient's age, 50 years. Comatose. The recording shows a mixture of symmetrical delta and theta with some spontaneous variability. His comatose state was secondary to abdominal sepsis. Calibration signal 1 s, 50 µV.

 

Fig. 6-6. Severe dysrhythmia. Patient's age, 7 years. Comatose. The first third of the illustration is dominated by beta, theta and delta (from benzodiazepines). The high voltage, slow frequency waves were triggered by vocal stimulus representing a pathological/encephalopathic alerting response. The patient had leukemia and had just recovered from a convulsive seizure several minutes before the recording. Calibration signal 1 s, 100 µV.

 

Fig. 6-7. Severe dysrhythmia. Patient's age, 67 years. Delirium. The recording shows a burst of rhythmic delta activity preceded and followed by runs of faster frequencies. The patient had hepatic encephalopathy. Calibrations signal 1 s, 70 µV.

 

Fig. 6-8. Severe dysrhythmia. Patient's age, 70 years. Comatose. The recording shows varied frequencies and amplitudes with delta and theta frequencies predominating. The patient suffered a gunshot wound to the left parietal lobe but the diffuse encephalopathy obscures any focal features. Calibration signal 2 s, 50 µV.

 

Fig. 6-9. Severe dysrhythmia. Patient's age, 53 years. Comatose. There is diffuse, invariant, semi-rhythmic delta with a paucity of faster frequencies. The patient had sepsis with endocarditis. He remains in a minimally conscious state with multiple cerebral infarctions. Calibration signal 1 s, 70 µV.

 

Fig. 6-10. No caption available.

 

Fig. 6-11. Triphasic waves. Patient's age 63 years. Coma. There is a burst of triphasic waves preceded by a mix of theta and delta followed by high voltage delta activity. Note the characteristic bilaterally synchronous burst with the prominent second phase. The patient had sepsis and multi-organ failure. Calibration signal 1 s, 70 µV.

 

Fig. 6-12. Triphasic waves. Patient's age, 58 years. Comatose. There are triphasic waves throughout the recording. Note the anterior-posterior shift of the major positive (down-going) wave of the triphasic complexes. The patient had pulmonary thromboembolism. Calibration signal 1 s, 70 µV.

 

Fig. 6-13. Triphasic waves. Patient's age, 76 years. Comatose. Triphasic waves are interrupted by or intermixed with rhythmic, frontally predominant delta activity, in itself characteristic of a metabolic encephalopathy when against a slow or suppressed background. The patient had hyperammonemia as an adverse effect of valproate therapy. Calibration signal 1 s, 100 µV.

 

Fig. 6-14. Triphasic waves. Patient's age, 90 years. Confused. Triphasic waves alternating with runs of theta and low voltage delta. Patient had pancytopenia and sepsis. Calibration signal 1 s, 150 µV.

 

Fig. 6-15. Triphasic waves versus eye blinks. Patient's age, 50 years. Comatose after cardiac arrest (no sedation). The top two channels are left and right infra-orbital leads (IOL, IOR) causing eye movement artifact (*) to “phase reverse” with the third and fourth channels. Triphasic waves occur periodically on a suppressed background. The patient did not recover consciousness. Calibration signal 1 s, 50 µV.

 

Fig. 6-16. No caption available.

 

Fig. 6-17. Burst-suppression pattern without epileptiform discharges. Patient's age, 7 years. Comatose. Drug-induced burst suppression pattern with bursts containing low-frequency theta. The patient was on propofol anesthesia for control of seizures. Calibration signal 2 s, 50 µV.

 

Fig. 6-18. Burst-suppression without epileptiform discharges. Patient's age, 80 years. Comatose. Burst suppression pattern with sharply contoured waves in the mid-posterior head which are not epileptiform and slow frequency waves. The patient suffered anoxic-ischemic encephalopathy from cardiac arrest and died without recovery of awareness. Calibration signal 1 s, 70 µV.

 

Fig. 6-19. Burst suppression without epileptiform discharges. Patient's age, 67 years. Profound coma, not sedated, with a profound burst-suppression pattern and low-voltage bursts. Calibration signal 2 s, 50 µV.

 

Fig. 6-20. Burst suppression with generalized epileptiform discharges. Patient's age, 38 years. Comatose. Intervening periods of suppression reveal ECG artifact. With anoxic–ischemic encephalopathy, such EEG patterns have a worse prognosis than burst-suppression patterns without epileptiform discharges (Synek, 1988). Calibration signal 1 s, 70 µV.

 

Fig. 6-21. Burst suppression with generalized epileptiform discharges. Patient's age, 46 years. Comatose. The burst suppression pattern shows bilaterally synchronous, frontally predominant spikes intermixed with theta and delta activity. The suppression periods show low voltage electrocardiographic artifact (note the inverse relationship of the main complexes between those channels with left and right ear references). The patient, who had chronic renal failure with superimposed pneumonia, had suffered a cardiac arrest with severe anoxic-ischemic encephalopathy. The patient died without recovering conscious awareness. Calibration signal 2 s, 70 µV.

 

Fig. 6-22. No caption available.

 

Fig. 6-23. Spindle coma. Patient's age, 15 years. Comatose with severe closed head injury. Also on midazolam infusion. Calibration signal 1 s, 50 µV.

 

Fig. 6-24. Spindle coma. Patient's age, 20 years. Comatose. The recording shows variable amplitudes and frequencies, but some approach a “spindle envelope” as in the mid and last half of this illustration. The patient had meningococcal septicaemia, but made a favourable recovery. Calibration signal 1 s, 70 µV.

 

Fig. 6-25. Alpha–theta pattern. Patient's age, 87 years. Comatose. There is a mixture of alpha, theta and beta frequencies. The patient had severe anoxic-ischemic encephalopathy and died without recovering conscious awareness. Calibration signal 1 s, 50 µV.

 

Fig. 6-26. Alpha coma. Patient's age, 43 years. Comatose. The EEG shows an alpha pattern coma with the alpha frequency highest in the anterior head and invariant. There are theta and delta frequencies in the posterior head as well, the patient was in coma from cardiac arrest and died without recovering consciousness. Calibration signal 1 s, 30 µV.

 

Fig. 6-27. No caption available.

 

Fig. 6-28. Focal spikes. Patient's age, 67 years. Comatose. The recording shows mild suppression with low voltage delta and minimal theta activity. There are recurring interictal epileptiform discharges in the right hemisphere. The patient died without recovering consciousness after cardiac surgery and multiple strokes. The accompanying MRI scan is shown in Fig. 6-29. Calibration signal 1 s, 50 µV.

 

Fig. 6-29. MRI-Multiple embolic strokes and global ischemia related to cardiac surgery. Same patient as previous illustration. The diffusion-weighted MRI scan shows multiple (white) regions of cortical infarction. The patient was comatose and had seizures onsetting with leftward eye deviation. Postmortem examination showed multifocal cortical infarctions.

 

Fig. 6-30. Right hemisphere PLEDs plus and left hemisphere suppression. Patient's age, 43 years. Comatose. The EEG shows a PLEDs plus pattern in the right hemisphere and a burst suppression in the left hemisphere. The patient had a previous severe head injury and presented with status epilepticus. He eventually recovered back to baseline. Calibration signal 1 s, 100 µV.

 

Fig. 6-31. BIPLEDs-plus. Patient's age, 67 years. Comatose. Bilateral PLEDs-plus with multifocal spikes. Calibration signal 1 s, 50 µV.

 

Fig. 6-32. Continuous PLEDs after acoustic neuroma surgery. Patient's age, 63 years. Comatose. PLEDs in the left mid-posterior temporal–central–parietal regions are seen as repetitive sharply contoured waves. Calibration signal 1 s, 100 µV.

 

Fig. 6-33. Bilaterally synchronous epileptiform discharges. Patient's age, 56 years. Comatose. These discharges occur in both temporal lobes, higher on the right on this occasion. Note the spread to the frontal polar electrodes. Calibration signal 1 s, 50 µV.

 

Fig. 6-34. No caption available.

 

Fig. 6-35. Seizure. Onset is from channel 1 electrode (left frontal inferior temporal region) and channel 3 electrode (left temporal mastoid) as recorded with subhairline EEG (SHEEG). The right-sided electrodes (channels 2 and 4) are homologous to the left. There is abundant ECG artifact (*) as well. Seizure recorded with Datex four-channel subhairline montage (see following illustrations).

 

Fig. 6-36. Seizure (continued). The evolutionary changes of the seizure are more prominent on the left side (channels 1 and 3).

 

Fig. 6-37. Seizure, termination. Further evolution of the seizure and its termination can be appreciated, despite rather poor resolution and heavy contamination by ECG artifact.

 

Fig. 6-38. Focal seizure. Patient's age, 77 years. Comatose. There are quasi-periodic complexes (PLEDs) in the right posterior head (maximum P4 and T6) and a reduction of faster frequencies from the right cerebral hemisphere. The patient had suffered a right middle cerebral artery territory ischemic event (not completely infarcted, as shown by a perfusion-diffusion mismatch on MRI scanning). Calibration signal 1 s, 100 µV.

 

Fig. 6-39. Seizure(continued). Same patient as previous illustration. The PLEDs have evolved into polyphasic complexes (PLEDs plus pattern) and then into a seizure with rhythmic waves that vary in amplitude and frequency in the right posterior head. Calibration signal 1 s, 100 µV.

 

Fig. 6-40. Seizure (continued). Same patient as previous 2 illustrations showing evolutionary changes of the seizure in the right hemisphere with continuous changes of rhythmic waves intermixed with some spikes. Note that the seizure discharge is now involving the mid and even anterior parts of the right hemisphere and is likely inducing a similar rhythm in the left hemisphere. Calibration signal 1 s, 100 µV.

 

Fig. 6-41. Seizure (continued). Same patient as in previous 3 illustrations. There is continued evolution of the seizure in the right hemisphere. The frequency has increased, but there is still a continued variability of the frequencies and amplitudes. Calibration signal 1 s, 100 µV.

 

Fig. 6-42. Seizure (continued). Same patient as previous 4 illustrations. The seizure is now interrupted by some slow frequency waves, especially in the right posterior head. Calibration signal 1 s, 100 µV.

 

Fig. 6-43. Seizure, (abating). Same patient as previous 5 illustrations. The seizure is showing waxing, then waning, then waxing of slow frequencies in the right hemisphere. Calibration signal 1 s, 100 µV.

 

Fig. 6-44. Seizure (termination). Same patient as previous 6 illustrations. The seizure has further evolved into spikes in the right superior frontal-central region intermixed with slower frequencies. The amplitude of spikes diminishes just before the electrographic seizure ends with right hemispheric suppression in the last 1.5 seconds of this illustration. Calibration signal 1 s, 100 µV.

 

Fig. 6-45. Generalized periodic epileptiform discharges (GPEDs or GPDs). Patients age, 31 years. Comatose. Generalized, roughly periodic complexes consisting chiefly of bilaterally synchronous single or repetitive spikes against a suppressed background. The patient had suffered severe anoxic-ischemic encephalopathy with myoclonic status epilepticus. Despite valproate and then propofol therapy, the patient died without recovering conscious awareness. Calibration signal 1 s, 150 µV.

 

Fig. 6-46. GPEDs. Patient's age, 75 years. Comatose. Widely spaced periodic generalized complexes, usually containing definite epileptiform discharges against a suppressed background. The patient suffered severe anoxic-ischemic cerebral cortical damage from cardiac arrest and did not have myoclonus. He died without recovering conscious awareness. Calibration signal 1 s, 70 µV.

 

Fig. 6-47. GPEDs. Patient's age, 80 years. Comatose. Triphasic-like waves, but these have a more prominent initial negative component that often exceeds the second (positive) component in amplitude. The background is slow. The patient had suffered a cardiac arrest and was unresponsive afterwards. Calibration signal 1 s, 70 µV.

 

Fig. 6-48. GPEDs. Same recording as previous illustration. Note the GPEDs have a topography that would be atypical for triphasic waves, as the latter usually show a more striking predominance of the frontal polar superior frontal derivations. Calibration signal 1 s, 100 µV.

 

Fig. 6-49. GPEDs resembling triphasic waves (continued). Same patient as previous 2 illustrations. These waves evolve into a series of sharply contoured waves that are often notched. This probably represents an electrographic seizure. Calibration signal 1 s, 100 µV.

 

Fig. 6-50. Further evolution of ictal discharge, maximum anterior head. Same patient as previous 3 illustrations. There is a further evolution of the electrographic (non-convulsive) seizure that changes to rhythmic delta and theta in the last second of this illustration. Calibration signal 1 s, 100 µV.

 

Fig. 6-51. No caption available.

 

Fig. 6-52. Lateralized suppression. Patient's age, 3 years. Confused. The recording shows a relative suppression of voltage in the entire right hemisphere. The left hemisphere shows polymorphic delta and occasional spikes from the posterior temporal region. The patient was admitted with focal left hemisphere status epilepticus. Calibrations signal 1 s, 150 µV.

 

Fig. 6-53. Complete suppression. Patient's age, 6 years. Comatose. The EEG is completely suppressed and contains only pulse and EMG artifact. The patient suffered anoxic-ischemic encephalopathy and died without regaining consciousness. Calibration signal 1 s, 50 µV.

 

Fig. 6-54. Incomplete suppression. Patient's age, 69 years. Comatose. This very low voltage tracing contains a poorly sustained mix of delta, theta, alpha and beta frequencies. The patient had Lewy body dementia and was involved in a motor vehicle accident. He suffered a traumatic brain injury with bilateral subdural hematomas. After neurosurgical drainage, he eventually recovered back to his very impaired baseline. Calibration signal 1 s, 70 µV.

 

Fig. 6-55. Incomplete suppression. Patient's age, 77 years. Comatose. The posterior head is suppressed, while there is discontinuous, arrhythmic, low voltage delta in the anterior head and superimposed muscle artifact. The patient had suffered a cardiac arrest and had absent brainstem reflexes at the time of the EEG (2 days post arrest). He died without recovering consciousness. Calibration signal 1 s, 50 µV.

 

Fig. 6-56. Suppression. Patient's age, 4 years. Comatose. All is artifact: pulse, electrode, EMG and electroretinogram (note the flash marking on the bottom). Calibration signal 1 s, 30 µV.

 

Fig. 6-57. Suppression with pulse artifact. Patient's age, 4 years. Comatose with sedation for seizures. Note the rate of the “rhythmic delta” in channel 1 is the same frequency as the recorded EKG. Calibration signal 1 s, 50 µV.

 

Fig. 6-58. No caption available.

 

Fig. 6-59. Focal arrhythmic delta. Patient's age, 17 months. Comatose. There is high voltage, arrhythmic delta in the left hemisphere, most marked posteriorly. There is also an asymmetry of faster frequencies, likely due to the skull defect on the left. The child presented with refractory focal motor seizures after drainage of a large left subdural hematoma. Calibration signal 1 s, 100 µV.

 

Fig. 6-60. CT scan. Same patient as previous illustration. Acute subdural hematoma over the left hemisphere. Note shift of midline structures from left to right.

 

Fig. 6-61. Lateralized delta and suppression of faster frequencies. Patient's age, 45 years. The right hemisphere shows arrhythmic delta that is most marked in the anterior head. Note the loss of faster frequencies from the right hemisphere compared to the left. The patient had a subarachnoid hemorrhage from a ruptured aneurysm, then suffered infarction of the right hemisphere from vasospasm. Calibration signal 1 s, 100 µV.

 

Fig. 6-62. CT scan. Same patient as previous illustration. There is acute infarction (with mass effect) in the right middle and anterior cerebral artery territories (left side of the scan).

 

Fig. 6-63. Severe left hemispheric suppression and burst suppression in right hemisphere. Patient's age, 20 months. Severe head trauma. The child was sedated with thiopental, likely producing the burst suppression in the less damaged right hemisphere. Calibration signal 1 s, 50 µV.

 

Fig. 6-64. Spindle and beta asymmetry. Patient's age, 35 years. Comatose. Note the relative suppression of beta in the right frontal region compared to the left. Delta activity is most prominent in the left anterior mid-temporal and central regions. Patient had a severe traumatic brain injury (no skull defect) and post-traumatic epilepsy. Medication included beta-producing clonazepam. Calibration signal 1 s, 50 µV.

 

Fig. 6-65. Asymmetry of beta. Patient's age, 32 years. Sleep. There is a relative reduction of beta and spindles from the left hemisphere. This patient had a large left congenital porencephalic cyst and seizure disorder. He was on beta-producing clobazam at the time of the recording. Calibration signal 1 s, 50 µV.

References

Al Thenayan E, Savard M, Sharpe M, et al. EEG for prognosis after cardiac arrest. J Clin Neurophysiol. In press.

Blume WT, Young GB, Lemieux JF. EEG morphology of partial epileptic seizures. Electroencephalogr Clin Neurophysiol. 1984;57:295–302.

Boulanger DM, Deacon C, Lecuyer D, et al. Triphasic waves vs. nonconvulsive status epilepticus: EEG distinction. Can J Neurol Sci. 2006;33:175–180.

Bridgers SL, Ebersole JS. EEG outside the hairline: Detection of epileptiform abnormalities. Neurology. 1988;38:146–149.

Claassen J, Meyer SA, Kowalski RG, et al. Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology. 2004;62: 1743–1748.

De Giorgio CM. The value of the EEG in determining brain death. Arch Neurol. 1989;46:602–603.

Fischer C. The use of EEG in the diagnosis of brain death in France. Neurophysiol Clin. 1997;27:373–382.

Fishgold H, Mathis P. Obnubilations comas et stupeurs: études electroencephalographiques. Electroencephalogr Clin Neurophysiol. 1959;11(Suppl. 11):27–68.

Hansotia P, Gottschalk P, Green P, et al. Spindle coma: Incidence, clinicopathologic correlates and prognostic value. Neurology. 1981;31:83–87.

Hebb MO, McArthur DL, Alger J, et al. Impaired percent alpha variability on continuous electroencephalography is associated with thalamic injury and predicts poor long-term outcome after human traumatic brain injury. J Neurotrauma. 2007;24:579–590.

Hirsch LJ, Brenner RP, Drislane FW, et al. The ACNS subcommittee on research terminology for continuous EEG monitoring: Proposed standardized terminology for rhythmic and periodic EEG patterns encountered in critically ill patients. J Clin Neurophysiol. 2005;22:128–35.

Jordan KG. Neurophysiologic monitoring in the neuroscience intensive care unit. J Clin Neurophysiol. 1995;13:579–626.

Moshé SL. Usefulness of EEG in the evaluation of brain death in children: The pros. Electroencephalogr Clin Neurophysiol. 1989;73:272–275.

Nairismagi J, Grohn OH, Kettunen MI, et al. Progression of brain damage after status epilepticus and its association with epileptogenesis: A quantitative MRI study in a rat model of temporal lobe epilepsy. Epilepsia. 2004;45: 1024–1034.

Reiher J, Rivest J, Grand'Maison F, et al. Periodic lateralized epileptiform discharges with transitional rhythmic discharges: Association with seizures. Electroencephalogr Clin Neurophysiol. 1991;78:12–17.

Savard M, Thenayan EA, Norton L, et al. Continuous EEG monitoring in severe Guillain–Barré syndrome patients. J Clin Neurophysiol. 2009;26:21–23.

Schneider S. Usefulness of EEG in the evaluation of brain death in children: The cons. Electroencephalogr Clin Neurophysiol. 1989;73:276–278.

P.589

 

Sediri H, Bourriez JL, Derambure P. Role of EEG in the diagnosis of brain death. Rev Neurol. 2007;163:248–353.

Seet RCS, Lim ECH, Wilder-Smith EPV. Spindle coma from acute midbrain infarction. Neurology. 2005;64:2159–2160.

Sundaram MB, Blume WT. Triphasic waves: Clinical correlates and morphology. Can J Neurol Sci. 1987;14:136–140.

Synek VM. Prognostically important coma patterns in diffuse anoxic and traumatic encephalopathies. J Clin Neurophysiol. 1988;5:161–174.

Vespa PM, Neno VV, Newer MR. Continuous EEG monitoring in the intensive care unit: Early findings and clinical efficacy. J Clin Neurophysiol. 1999; 16:1–13.

Westmoreland BF, Klass DW, Sharbrough FW, et al. Alpha coma. Electroencepalographic, clinical, pathologic and etiologic considerations. Arch Neurol. 1974;24: 582–588.

Young GB. The EEG in coma. J Clin Neurophysiol. 2000;17(5):473–485.

Young GB, Blume WT, Campbell VM, et al. Alpha, theta and alpha–theta coma: A clinical outcome study using serial recordings. Electroencephalogr Clin Neurophysiol. 1994;91:93–99.

Young GB, Bolton CF, Austin TW, et al. The electroencephalogram in sepsis-associated encephalopathy. J Clin Neurophysiol. 1992;145–152.

Young GB, Campbell VC. EEG monitoring in the intensive care unit: Pitfalls and caveats. J Clin Neurophysiol. 1999;16:40–45.

Young GB, Doig GS. Continuous EEG monitoring in comatose intensive care unit patients: Epileptiform activity in etiologically distinct groups. Neurocrit Care. 2005;2:5–10.

Young GB, Jordan KG. Do nonconvulsive seizures damage the brain? Yes. Arch Neurol. 1998;55:1117–1119.

Young GB, Jordan KG, Doig GS. An assessment of nonconvulsive seizures in the intensive care unit using continuous EEG monitoring: An investigation of variables associated with mortality. Neurology 1996;47:83–89.

Young GB, Kreeft JH, McLachlan RS, et al. EEG and clinical associations with mortality in comatose patients in a general intensive care unit. J Clin Neurophysiol. 1999;16:354–360.

Young GB, Lee D. A critique of ancillary tests for brain death. Neurocrit. Care 2004; 1:499–508.

Young GB, McLachlan RS, Kreeft JH, et al. An electroencephalographic classification system for coma. Can J Neurol Sci. 1997;24:320–325.

Young GB, Sharpe MD, Savard M, et al. Seizure detection with a commercially available bedside EEG monitor and the subhairline montage. Neurocrit Care. 2009; July 8 [Epub ahead of print.]