Other Biomedical Signals
241
The main processing techniques applicable to EOG are FT and wavelet transform
(WT). Certain domain-related features are also calculated from EOG. For example,
the difference between the timing of the cursor move and the eye response can be
calculated simply based on the direct measurement of the time difference between
the rising edges of the two signals.
The long recordings of EOG are also used for the study of sleep. Figure 12.4
shows the recordings of the right and left EOGs, sampled at 50 Hz during a sleep
study in which a number of physiological signals during sleep are recorded and
correlated.
A very closely related signal called “electroretinogram” or ERG has been used
for very similar applications. This signal that is the potential difference among the
retina and the surface of the eyeball is known to be highly correlated with EOG, and,
as a result, EOG is often used in many applications to represent ERG.
12.4 MAGNETOENCEPHALOGRAM
The signal known as “magnetoencephalogram,” or MEG, is essentially the magnetic
equivalent of EEG. In other words, while EEG captures the activities of the brain
neurons through detection of the changes in the electric activities on the surface of
the head, MEG measures the changes in the magnetic field caused by the activities of
the brain neurons. From electromagnetics, we know that any changing electric field
causes a magnetic field that is proportionally related to the electric field. As a result,
the changes in the electric charges of the neurons create a magnetic field that can be
measured to detect the activities of the brain. MEG measures the extracranial magnetic fields produced by intraneuronal ionic current flow within appropriately oriented
cortical pyramidal cells.
At this point, we need to address the following question: “If MEG captures almost
the same information as EOG, why do we need MEG at all?” The question becomes
more relevant if we consider the fact that the instrumentation needed to measure MEG
is significantly more complex and expensive than that of EEG. It seems that since we can
capture the same information using much less expensive EEG machines, there would be
no need for MEG. The answer to this question is twofold. First, EEG is captures on the
surface of the skull and therefore is suitable to many sources of noise such as the electric
activities of the muscles close to electrodes. The lack of skin contact facilitates using
MEG to record DC and very-high-frequency (>600 Hz) brain activity.
In addition, the MEG is capable of detecting the electric activities of the neurons
deeper in the brain, as opposed to the EEG signals that are often due to the neurons
closer to the surface of the brain. More specifically, MEG has selective sensitivity to
tangential currents (from fissural cortex) and less distorted signals compared with
EEG. This allows MEG to provide much better spatial and temporal accuracy. A
major advantage of MEG is determining the location and timing of cortical generators for event-related responses and spontaneous brain oscillations. MEG provides a
spatial accuracy of a few millimeters under optimal conditions, combined with an
accurate submillisecond temporal resolution, which together enable spatiotemporal
tracking of distributed neural activities, for example, during cognitive tasks or epileptic discharges.
241
The main processing techniques applicable to EOG are FT and wavelet transform
(WT). Certain domain-related features are also calculated from EOG. For example,
the difference between the timing of the cursor move and the eye response can be
calculated simply based on the direct measurement of the time difference between
the rising edges of the two signals.
The long recordings of EOG are also used for the study of sleep. Figure 12.4
shows the recordings of the right and left EOGs, sampled at 50 Hz during a sleep
study in which a number of physiological signals during sleep are recorded and
correlated.
A very closely related signal called “electroretinogram” or ERG has been used
for very similar applications. This signal that is the potential difference among the
retina and the surface of the eyeball is known to be highly correlated with EOG, and,
as a result, EOG is often used in many applications to represent ERG.
12.4 MAGNETOENCEPHALOGRAM
The signal known as “magnetoencephalogram,” or MEG, is essentially the magnetic
equivalent of EEG. In other words, while EEG captures the activities of the brain
neurons through detection of the changes in the electric activities on the surface of
the head, MEG measures the changes in the magnetic field caused by the activities of
the brain neurons. From electromagnetics, we know that any changing electric field
causes a magnetic field that is proportionally related to the electric field. As a result,
the changes in the electric charges of the neurons create a magnetic field that can be
measured to detect the activities of the brain. MEG measures the extracranial magnetic fields produced by intraneuronal ionic current flow within appropriately oriented
cortical pyramidal cells.
At this point, we need to address the following question: “If MEG captures almost
the same information as EOG, why do we need MEG at all?” The question becomes
more relevant if we consider the fact that the instrumentation needed to measure MEG
is significantly more complex and expensive than that of EEG. It seems that since we can
capture the same information using much less expensive EEG machines, there would be
no need for MEG. The answer to this question is twofold. First, EEG is captures on the
surface of the skull and therefore is suitable to many sources of noise such as the electric
activities of the muscles close to electrodes. The lack of skin contact facilitates using
MEG to record DC and very-high-frequency (>600 Hz) brain activity.
In addition, the MEG is capable of detecting the electric activities of the neurons
deeper in the brain, as opposed to the EEG signals that are often due to the neurons
closer to the surface of the brain. More specifically, MEG has selective sensitivity to
tangential currents (from fissural cortex) and less distorted signals compared with
EEG. This allows MEG to provide much better spatial and temporal accuracy. A
major advantage of MEG is determining the location and timing of cortical generators for event-related responses and spontaneous brain oscillations. MEG provides a
spatial accuracy of a few millimeters under optimal conditions, combined with an
accurate submillisecond temporal resolution, which together enable spatiotemporal
tracking of distributed neural activities, for example, during cognitive tasks or epileptic discharges.