Human electroencephalography (EEG) provides a convenient but often opaque "window on the mind,'' allowing observations of electrical processes near the brain surface. The outer brain layer is the cerebral cortex, believed to be largely responsible for our individual thoughts, emotions, and behavior. Cortical processes involve electrical signals that change over times in the 0.01-sec range. EEG is the only widely available technology with sufficient temporal resolution to follow these quick dynamic changes. On the other hand, EEG spatial resolution is poor relative to that of modern brain structural imaging methods—computed tomography, positron emission tomography, and magnetic resonance imaging (MRI). Each scalp electrode records electrical activity at large scales, measuring electric currents (or potentials) generated in cortical tissue containing approximately 30-500 million neurons.
Electrodes may be placed inside the skull to study nonhuman mammals or human epilepsy patients. Such intracranial recordings provide measures of cortical dynamics at several small scales, with the specific scale dependent on electrode size. Intracranial EEG is often uncorrelated or only weakly correlated with cognition and behavior. Human ''mind measures'' are more easily obtained at the large scale of scalp recordings. The technical and ethical limitations of human intracranial recording force emphasis on scalp recordings. Luckily, these large-scale estimates provide important measures of brain dysfunction for clinical work and cognition or behavior for basic scientific studies.
EEG monitors the state of consciousness of patients in clinical work or experimental subjects in basic research. Oscillations of scalp voltage provide a very limited but important part of the story of brain functioning. For example, states of deep sleep are associated with slower EEG oscillations of larger amplitude. More sophisticated signal analyses allow for identification of distinct sleep stages, depth of anesthesia, epileptic seizures, and connections to more detailed cognitive events. A summary of clinical and research EEG is provided in Fig. 1. The arrows indicate common relations between subfields. Numbers in the boxes indicate the following:
1. Physiologists record EEG from inside skulls of animals using electrodes with diameters ranging from approximately 0.001 to 1 mm. Observed dynamic behavior generally depends on measurement scale, determined by electrode size for intracranial recordings. In contrast, scalp-recorded EEG dynamics is exclusively large scale and mostly independent of electrode size.
2. Human spontaneous EEG occurs in the absence of specific sensory stimuli but may easily be altered by such stimuli.
3. Averaged evoked potentials (EPs) are associated with specific sensory stimuli, such as repeated light flashes, auditory tones, finger pressure, or mild electric shocks. They are typically recorded by time averaging to remove effects of spontaneous EEG.
4. Event-related potentials (ERPs) are recorded in the same way as EPs but occur at longer latencies from the stimuli and are associated more with endogenous brain states.
5. Because of ethical considerations, EEG recorded in brain depth or on the brain surface [electro-corticogram (ECoG)] of humans is limited to
patients, most of whom are candidates for epilepsy surgery.
6. With transient EPs or ERPs the stimuli consist of repeated short pulses. The number of pulses required to produce an average EP may range from approximately 10 to several thousand, depending on the application. The scalp response to each pulse is averaged over the individual pulses. The EP or ERP in any experiment consist of a waveform containing a series of characteristic component waveforms, typically occurring less than 0.5 sec after presentation of each stimulus. The amplitude, latency from the stimulus, or covariance (in the case of multiple electrode sites) of each component may be studied, in connection with a cognitive task (ERP) or with no task (EP).
7. Steady-state EPs use a continuous sinusoidally modulated stimulus (e.g., a flickering light) typically superimposed in front of a TV monitor showing the cognitive task. The brain response in a narrow frequency band containing the stimulus frequency is measured. Magnitude, phase, and coherence (in the case of multiple electrode sites) may be related to different parts of the cognitive task.
8. Alzheimer's disease and other dementia typically cause substantial slowing of normal alpha rhythms. Traditional EEG has been of little use in dementia because EEG changes are often only evident late in the illness when other clinical signs are obvious. However, recent efforts to apply EEG to early detection of Alzheimer's disease have shown promise.
9. Cortical tumors that involve the white matter layer (just below neocortex) cause substantial low-frequency (delta) activity over the hemisphere with the tumor. Application of EEG to tumor diagnosis has been mostly replaced by MRI, which reveals structural abnormalities in tissue.
10. Most clinical work uses spontaneous EEG; however, multiple sclerosis and surgical monitoring are exceptions, often involving EPs.
11. Studies of sensory pathways involve early components of EPs (less than approximately 50 msec) since the transmission times for signals traveling between sense organ and brain are short com pared to the duration of multiple feedback associated with cognition.
12. The study of general intelligence, often associated with IQ tests, is controversial. However, many studies have reported substantial correlation between scores on written tests and different quantitative EEG measures.
13. Mathematical models of large-scale brain function are used to explain or predict observed properties of EEG in terms of basic physiology and anatomy. Although such models represent vast oversimplifications of genuine brain function, they contribute to a general conceptual framework and may guide the design of new experiments to test this framework.
Was this article helpful?