NeuroImage 89 (2014) 143–151
Contents lists available at ScienceDirect
NeuroImage
journal homepage: www.elsevier.com/locate/ynimg
Multi-channel atomic magnetometer for magnetoencephalography:
A configuration study
Kiwoong Kim a, Samo Begus b, Hui Xia c, Seung-Kyun Lee c, Vojko Jazbinsek d,
Zvonko Trontelj d, Michael V. Romalis c,⁎
a
Korea Research Institute of Standards and Science, South Korea
Faculty of Electrical Engineering, Ljubljana, Slovenia
Physics Department, Princeton University, NJ, USA
d
Institute of Mathematics, Physics and Mechanics, Ljubljana, Slovenia
b
c
a r t i c l e
i n f o
Article history:
Accepted 22 October 2013
Available online 1 November 2013
Keywords:
Atomic magnetometer
Auditory evoked field
Biomedical signal processing
Magnetoencephalography
a b s t r a c t
Atomic magnetometers are emerging as an alternative to SQUID magnetometers for detection of biological magnetic fields. They have been used to measure both the magnetocardiography (MCG) and magnetoencephalography (MEG) signals. One of the virtues of the atomic magnetometers is their ability to operate as a multi-channel
detector while using many common elements. Here we study two configurations of such a multi-channel atomic
magnetometer optimized for MEG detection. We describe measurements of auditory evoked fields (AEF) from a
human brain as well as localization of dipolar phantoms and auditory evoked fields. A clear N100m peak in AEF
was observed with a signal-to-noise ratio of higher than 10 after averaging of 250 stimuli. Currently the intrinsic
magnetic noise level is 4 fTHz−1/2 at 10 Hz. We compare the performance of the two systems in regards to current source localization and discuss future development of atomic MEG systems.
© 2013 Elsevier Inc. All rights reserved.
Introduction
One of the most successful applications of the superconducting
quantum interference device (SQUID) is in the field of biomagnetism.
Especially, high sensitivity of a low-Tc SQUID magnetometer enabled
the measurement of neuromagnetic fields from a human brain and
opened the field of magnetoencephalography (MEG) (Cohen, 1972).
Since the SQUID MEG system was developed, numerous successful
investigations in various fields have been conducted with MEG. At
present, MEG is one of the most useful modalities for studies of brain
functions together with the functional magnetic resonance imaging
(fMRI) and positron emission tomography (PET). Specifically, due to
its high temporal resolution, MEG has a unique position in the field of
the cognitive science (Kwon et al., 2005; Tarkiainen et al., 2003). Although MEG is a powerful tool for functional brain imaging in both
basic brain research (Darvas et al., 2003; Kakigi et al., 1995) and clinical
diagnoses (Cheyne et al., 2007; Colon et al., 2009), existing SQUID systems have a number of technical limitations that hinder more widespread use of this technique. Among them are a high rate of liquid
helium consumption, fixed sensor configuration, frequent requirement
for cryogenic maintenance, high cost and the need for large shielded
rooms.
⁎ Corresponding author.
E-mail address: [email protected] (M.V. Romalis).
URL: http://physics.princeton.edu/atomic/romalis/ (M.V. Romalis).
1053-8119/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.neuroimage.2013.10.040
Atomic magnetometer (AM) which optically detects polarization
change of alkali metal vapor under external magnetic field is emerging
as a promising alternative to existing MEG sensors. The sensitivity of a
spin-exchange relaxation free (SERF) magnetometer is sufficient for
measuring an MEG signal (Kominis et al., 2003). In addition to the
absence of cryogenics, atomic magnetometers can operate in a multichannel configuration at a relatively low cost by utilizing many common
detector elements. They can also operate with a smaller magnetic shield
since there is no need to place a bulky liquid helium dewar inside the
shielded room. However, the geometrical constraints and other technical features of the AM system are substantially different from SQUID
systems and require careful consideration for maximum utilization
of the available sensitivity and for being practically useful in routine
MEG studies.
The aim of this research is to show the ability of potassium AM as a
detector of weak evoked brain signals. In this study, we compare specific features of two multi-channel AM MEG configurations and introduce
AM specific data analysis methods such as localization of effective sensor positions by application of gradients, noise reduction technique for
mode hopping elimination (MHE), and response time correction. The
first system uses a transmitted probe beam, similar to the arrangement
reported in Xia et al. (2006). In the second system we use a retroreflected probe beam and several new technical features. We show
measurements of auditory evoked fields (AEF) with both AM systems
and demonstrate the source localization with dipolar phantoms and
AEF signals.
144
K. Kim et al. / NeuroImage 89 (2014) 143–151
In addition to our approach using a single alkali-metal cell for multichannel recording, it is also possible to use an array of individual fibercoupled atomic sensors. Fiber-coupled sensors with small alkali-metal
cells have reached sensitivity below 10fT/Hz1/2 (Shah and Romalis,
2009) and have been used for recording of magnetoencephalography
signals (Johnson et al., 2010, 2013; Sander et al., 2012).
Material and methods
Transmitted probe atomic MEG system
Generals
The detailed description of this setup was reported previously (Xia
et al., 2006). By absorbing circularly polarized pumping photons, potassium atoms in a glass cell are spin-polarized, creating local magnetization.
The magnetization is rotated in the presence of an external magnetic
field. The tipped component of magnetization rotates the polarization
of linearly polarized probe light. We optically measure the distribution
of By field components in the y–z plane by using a 16 × 16-channel photodetector array, with geometry shown in Fig. 1(a).
In comparison to SQUID systems, where the sensor positions are
fixed at design state, the AM system has more flexibility. In our case
the head of a subject is placed on top of the cell, as shown in Fig. 2.
This feature is advantageous, for example, for baby MEG detection
(Okada et al., 2006). An inherent restriction in the optical signal detection is the fact that we lose the position information along the probe
beam direction; i.e. what we measure is the integral of all the magnetic
fields along the probe beam path. However, to localize a neuroelectric
source, we need to also get the spatial field distribution along the x
axis. This can be achieved by slicing the pumping beam in multiple sections and illuminating only part of the cell at any given time (Gusarov
et al., 2009). Such time division measurement is possible for repetitive
measurements, such as evoked fields. In the detection of the spontaneous brain wave or epileptic spikes, one can also use magnetic field modulation techniques to measure different components of the field (Li,
2006; Seltzer and Romalis, 2004).
In our configuration, the sensor arrangement also provides a depth
profile of the magnetic field (Fig. 1(c)), which can help with magnetic
source multipole analysis. The depth profile measures the decrease of
the magnetic field as a function of distance from the source to a sensing
location. The decay of the magnetic fields depends on the order of
the current multipole, hence one can analyze the source current structure, particularly for the case of multiple sources. For a SQUID sensor
system, such layered configuration of pickup coils is undesirable due
to the superconductive screening currents on the lower coils. Even
with adoption of an external feedback SQUID scheme, distortion of the
fields in the gradiometric pick-up coil configuration is unavoidable.
In contrast, high buffer gas pressure in the vapor cell of the atomic
Fig. 1. Multichannel magnetic field mapping with an atomic magnetometer; the expanded probe laser beam detects the spatial distribution of By components on the y–z plane with a 256channel detector array. (b) Spatial distribution of pumping rate in the vapor cell, which makes a different sensor characteristic for each sensor. (c) Map of the magnetic field map of the AEF
signal measured with the magnetometer.
K. Kim et al. / NeuroImage 89 (2014) 143–151
145
for low-pass response of the magnetometer. We note that even though
the response of atomic magnetometers drops at higher frequencies due
to limited bandwidth, the noise also often drops, so the sensitivity of the
magnetometer can be flat over a frequency range substantially larger
than its bandwidth (Shah et al., 2010).
Mode hopping elimination
Mode hopping in the semiconductor laser generates big stepwise
jumps in the signal baseline. Even averaging hundreds of the response
epochs is not enough to suppress this noise. By doing singular value decomposition on the covariance matrix of measured fields without stimulus, we can separate the large variance noise eigen-components and
their corresponding spatial field patterns (Fig. 3). Elimination of the projection components of these spatial patterns effectively reduces not
only the mode hopping noise, but also other ambient magnetic field
noises and possibly spontaneous brain magnetic fields.
Retro-reflected probe atomic MEG system
Fig. 2. Auditory evoked field measurement with the atomic magnetometer system. The
potassium cell and a human subject are placed in a three-layer cylindrical Mu-metal shield
with inner diameter of 0.9 m and outer diameter of 1 m. To block heat flow from the oven
containing the cell, chilled water is circulating through a thin water bag between the head
and the potassium cell's heating system. Tone stimuli are applied to one ear through a
non-magnetic pneumatic earphone.
magnetometer reduces the cross-talk between the sensing channels
and we can measure nearly continuous spatial field distributions. The
cross-talk free distance lD is determined by the distance that Rb atoms
can diffuse during their spin coherence time. This distance is proportional to the square root of the ratio of the diffusion constant D to the
spin relaxation rate Γ:
rffiffiffiffi
D
lD ¼
:
Γ
ð1Þ
For our conditions the diffusion constant of Rb atoms in 1 atm. of He
buffer gas at 180 °C is 0.45 cm2/s. The total spin relaxation rate of Rb
atoms Γ is equal to 2π × 30 Hz, which is the bandwidth of the magnetic
field response. The diffusion distance is then equal to 0.5 mm, much
smaller than the actual separation of the magnetometer channels.
Therefore, each channel can be considered to be independent.
Multichannel characteristics
In optical polarimetry, the signal is a product of light intensity and
the rotation of light polarization. As each channel is illuminated with
slightly different probe light intensity, we need to calibrate each channel with a known magnetic field. Moreover, due to the finite optical
depth, each sensor has a different pumping rate along the pumping
beam direction (Fig. 1(b)). The sensitivity to By depends on the
pumping rate, and the magnetometer bandwidth also depends on the
optical pumping rate as well as the polarization of atoms (Kominis
et al., 2003). Therefore, spatially inhomogeneous pumping profile produces different characteristics of each channel both in sensitivity and
in detection bandwidth. If the frequency range of a signal of interest is
outside of the sensor bandwidth, one can get a significant phase delay
of the signal component. Event related potential usually forms a peak;
the delay will change the latency of the peak appearance, necessitating
appropriate correction, such as a digital high-pass filter to compensate
As discussed in the previous sections, the source localization capability is one of the most important design factors of an atomic magnetometer system for the biomagnetism applications. The problems
mentioned above were dealt with in our new system design (Fig. 4).
First, we adopted a retro-reflected probe beam for circumventing
the blind direction. At the mirror placed on the measurement surface,
the probe beam reflects back to the detector after passing through the
potassium cell twice. The rotation angle of the probe beam polarization
due to paramagnetic Faraday rotation is added after two passes. The
reflected beam is recorded by 16 × 16 channel photodiode array and
we can measure the spatial distributions of By and Bz components by alternating the pumping beam direction with the beam switch as shown
in Fig. 4(b). The By and Bz components are corresponding to the orthogonal tangential field components of the MEG. Compared to the radial
component (Bx) measurement, the tangential components measurement provides deeper and wider localizable space for current dipole
sources when we have the same detection coverage areas (Kim et al.,
2004). In the tangential component measurement, the field maximum
pattern will be located right over the current dipole source while the radial component measurement shows a diverse dipolar pattern. Besides
the sensor coverage area, the By and Bz components guarantee the information orthogonality even for spatially correlated fields at nearby sensors; this configuration carries more information on the sources.
Second, we use two separate pumping lasers which are detuned in the
opposite direction from the D1 line in K. This reduces their absorption
cross-section and allows the light to propagate further into the cell to
obtain a more uniform spatial pumping profile. The detuning of the
Fig. 3. Spatial probe beam profile corresponding to the mode hopping of the laser diode.
146
K. Kim et al. / NeuroImage 89 (2014) 143–151
Fig. 4. Schematic diagram for the retro-reflected probe atomic MEG system. (a) The widely-expanded probe beam reflects back to the photodiode array. In the wide K vapor cell
(12 × 12 × 4 cm), the beam path is doubled. It measures the tangential field distribution from the brain. (b) By alternately switching the beam switch, we can change the direction of
the pumping beam to measure two orthogonal tangential field components, By and Bz. The balanced detuning of the pump beam compensates the light shift of each beam and the detuned
beam provides more uniform pumping profile.
two lasers from resonance is equal and opposite in order to cancel the
virtual magnetic field, so-called light shift, created by laser light detuned
from the optical resonance in K atoms. The light is generated by two DFB
lasers which are combined and amplified by a 1-W tapered amplifier to
supply high enough power for pumping the 12-cm deep vapor cell
(Souza, 2008). Third, the heating system has been changed from hot
air flow to AC electric heating. A pair of non-magnetic heating wires
was twisted to prevent from generating magnetic fields in the cell.
The heating power is generated by an audio amplifier at 20 kHz and is
far away from the detection bandwidth. The operation temperature
was about 200 °C to get high enough potassium density.
Results
Transmitted probe system
N100m peak is a typical primary brain response that appears
100 ms after hearing a single tone stimulus. Human subject recordings
followed a protocol approved by Princeton University Institutional Review Board. Since the main focus of this research is on the experimental
apparatus for acquiring MEG data, we used just a few subjects as representative examples of typical MEG signals. We applied 500-Hz tone
stimuli to the subject's right ear using a pneumatic earphone and
measured the response of the contra-lateral cortex. A clear N100m
peak in AEF was observed with a signal-to-noise ratio of higher
than 10 after averaging of 250 stimuli (Fig. 5(a)). For comparison,
Fig. 5(b) shows 100-times-averaged AEF measured by a hemispherical
37-channel double relaxation oscillation SQUID (DROS) magnetometer
system having a 3 fTHz−1/2 white noise level (Lee et al., 2003). In terms
of sensitivity, the two modalities are comparable to each other. The single polarity peaks in the AM recording imply that the current AM system does not have enough detection area (~3 cm in a length) to cover
the dipolar pattern of the magnetic fields, while the hemispherical
SQUID system covers an area of 14 cm in diameter.
As a rule of thumb, to minimizepthe
ffiffiffi localization error of a current dipole source, we require at least a 2 times wider measuring diameter
than the distance between the sensor array and the source when the
source is placed
at the center of the array. Once we define the measuring
pffiffiffi
diameter as L, 2 times the depth of the target source current dipole, all
the lengths can be normalized by L. We calculated the required signal to
noise ratio for the source localization as functions of the coverage
diameter and the deviation of the center of the detection area from
the position of a target current dipole source, respectively (Fig. 6). The
calculation result shows the minimum required SNR with which 95%
of current dipoles could be localized in 1 cm3 error volume. As shown
in Fig. 6, for a MEG source on the brain cortex, the distance between
K. Kim et al. / NeuroImage 89 (2014) 143–151
147
Fig. 5. (a) Auditory evoked field traces for all atomic magnetometer channels. The typical N100m peak appears 100 ms after the sound stimulus. (b) AEF traces measured by a 37-channel
SQUID MEG system.
the sensor array and the source is about 60 mm, and L ~ 85 mm. The required diameter of the detection area, 1.4L, is about 12 cm. Since the
exact source location is not known, the cell size should be even larger
to ensure optimal localization.
Another issue with the AEF localization using an AM is the direction
of the source dipole. In our configuration, the AEF current dipole happened to be oriented nearly perpendicular to the probe beam direction.
As mentioned in the previous section, the probe beam integrates all the
field distribution along the beam path. If the current dipole was located
in the middle of the sensing area, the positive and the negative fields
from the dipole would cancel out each other and the signal amplitude
would be reduced. This problem can be solved with slicing the pumping
beam. The optimal number of slices was found by using a computer simulation minimizing the source localization error. Interestingly, slicing
the pump beam in two gave the lowest localization error (Fig. 7). In
our configuration, the higher number of slices gives the lower signal intensity, which results in a larger localization error. We can realize
the pumping slice selection by using alternative half-blocking of the
pumping beam. To investigate this issue further, we placed a current dipole around the middle of the sensing area by using a spherical head
conductor phantom consisting of a current dipole in saline solution.
We applied currents simulating the AEF and alternatively measured
each pumping area; we alternatively blocked each half of the pumping
beam cross-section. Fig. 8 shows the measurement results. Figs. 8(a)
and (c) are the measured waveforms of all the channels while the first
half of the pump beam is blocked and the spatial field distribution at
the instant of the maximum peak, respectively. Figs. 8(b) and (d) are
the measured waveforms of all the channels while the other half of
the pump beam is blocked and the spatial field distribution at the instant of the maximum peak, respectively. With these two sets of the
spatial magnetic field distribution, we could make source dipole localization by solving an iterative nonlinear optimization problem with variables for three orthogonal components of the position and the two
components of the current dipole strength; the spherical conductor
model eliminates the other source component of the dipole (Sarvas,
1987). The estimated position of the current dipole was deviated from
the original position by Δx ~ 15 mm, Δy ~ 1 mm, Δz ~ 1 mm. Of
course, the error in the direction for the probe beam was largest.
Retro-reflected probe atomic MEG system
Prior to measurement of the audio evoked brain signals with AM a
compensation of the residual magnetic fields in the volume of the potassium cell is necessary. We used the compensation coils inside the inner
magnetic shield. This way the SERF operation of the AM was assured.
The AM channels sensitivity was calibrated by applying low frequency
magnetic field (50 pT, 10 Hz) across the AM cell. Noise spectrum of
one magnetometer's channel and of the two adjacent channels in the
gradiometer configuration is shown in Fig. 9. The intrinsic magnetometer noise determined by dividing thepnoise
in the difference between
ffiffiffi
two adjacent channels by a factor of 2 is 4 fTHz−1/2 above 10 Hz. At
higher frequencies, vibration and magnetic interference peaks are significant, but they are partly canceled in the gradiometer measurements.
By applying low frequency constant magnetic field gradient
(10 pT/cm, 10 Hz) using the gradient compensation coils the spatial
mapping information from the AM cell to each photodiode position
was obtained. The two-dimensional field profile data were smoothed
10
Localization error volume, cm
3
100 nAm dipole || z−axis
100 nAm dipole || y−axis
8
6
4
2
0
Fig. 6. Required signal-to-noise ratio for the source localization as functions of source position deviation from the center of the p
cell
ffiffiffi and detection area, respectively. The L is defined
to be a measuring diameter which is 2 times the depth of the target source current dipole. The detection area over 1.4L shows a flat SNR profile less depending on the position
of the source.
1
2
4
8
16
Number of slices (pumping selections)
Fig. 7. Simulated source localization error as a function of the number of the pumping
slices.
148
K. Kim et al. / NeuroImage 89 (2014) 143–151
Fig. 8. Results of the sliced pumping experiment for the current dipole localization with a spherical brain phantom. One half of the pump beam is blocked alternatively. (a) The measured
waveforms of all the channels while the first half of the pump beam is blocked. (b) The measured waveforms of all the channels while the other half of the pump beam is blocked (c) and
(d) are the spatial field distributions at the instant of the maximum peaks of (a) and (b), respectively.
with an average value of 5 by 5 elements (Reeves, 2009) before the spatial information has been calculated (Fig. 10).
Effective calculated positions of the magnetometer channels using
the magnetic field gradient mapping are shown in Fig. 11. Channels
with good signal to noise ratios, used in data analysis are shown with
large marks.
4
10
Ch133
Ch133 − Ch188
3
B, fTHz
−1/2
10
2
10
1
10
0
10
10
20
30
40
50
f, Hz
Fig. 9. Noise spectrum of single channel magnetometer and gradiometer configuration of
the two adjacent channels.
Each auditory applied stimulus consists of 20 square pulses with
a period of 1 ms and duty cycle of 50%. The interval between each
pulse train is varied randomly between 1.3 s and 2.0 s to avoid subject's
adaptation.
Prior to data analysis the signals were filtered by a band-pass 2–20 Hz
filter. After rejecting the subject's heartbeat signals, signals originating
from eye movements and disturbances due to mechanical vibrations,
the N100m could be seen in several channels of the 256 channel AM
after averaging 710 stimulus epochs. This was achieved by combining
By magnetometer channels into a gradiometer configuration: one magnetometer was selected as a reference channel and was subtracted from
the other channels. Fig. 12 shows gradiometric channels with the best
signal to noise ratio.
Analysis of MEG data followed common biomagnetism techniques.
We have chosen a current dipole source model in a conducting sphere
(Sarvas, 1987). Using signals at 100 ms after the stimulus from 10 selected By gradiometric channels with good signal to noise ratio we
found the position (rp), the dipole moment p = (p2x + p2y)1/2 and the
angle ϕ = arctan(px/py) for the best-fitting equivalent current dipole.
The third dipole component pz was neglected in the analysis because
in our setup it was almost radial and therefore did not contribute significantly to the magnetic field outside of the conducting sphere (Sarvas,
1987). Relative error (RE) defined as a root mean square (RMS) difference between measured and calculated data divided by the RMS of
measured data was 0.1 and corresponding correlation coefficient (CC)
0.99. The obtained location rp = (−6.36, 2.15, 0.34) cm relative to the
sphere center is in the expected region of the cortex (Pantev et al.,
K. Kim et al. / NeuroImage 89 (2014) 143–151
(b)
8
10
5
15
4
−40
2
0
15
0
−25
−30
5
−30
35
40
30
35 30
25
20
15
10
5
0
−5
−10
20
10
0
−10
−20
15
−5
6
30
25
20
15
10
5
0
−5
−10
−15
−10
0
0
35
25
20
15
10
5
0
15
10
5
2
10
5
0
−5
−10
−15
−20
4
40
30
−
0
20
15
12
40
B, pT
0
5
10
15
B, pT
−35
−30
−25
−20
−15
−
−5 10
−10
−20
−4
14
0
5
6
0 5
10
15
10
−3
8
−45
10
−40
Row nr.
12
16
10
25
−30
14
−30
−35
Row nr.
16
−25
−20
−15
−10
−5
(a)
149
−20
0
5
Column nr.
10
15
Column nr.
Fig. 10. Smoothed measured magnetic field at different photodiode locations with applied constant magnetic field gradient over the potassium cell. Left: gradient in y-direction. Right:
gradient in z-direction.
1987), distance from the center (|rp| = 6.7 cm), the dipole moment
p = 74.1 nAm and the angle ϕ = 34.2°. Fig. 13 shows magnetic field
maps for the tangential and radial components calculated from the
equivalent current dipole.
The alternative way to solve the biomagnetic inverse problem is the
minimum norm estimation (MNE) method (Parker, 1977; Sarvas, 1987)
where only the source space is constrained. Among all possible current
distributions in that space, the one with the minimum norm is selected.
Fig. 14 shows reconstructed current distribution constrained to a spherical cap surface inside a conducting sphere. We have got the same RE
(0.1) and CC (0.99) as in the case of the equivalent current dipole
source. Fig. 15 shows tangential and radial components of magnetic
field calculated from the obtained MNE current distribution in Fig. 14.
Again, these magnetic field maps are very similar to those obtained by
a single equivalent dipole current source.
These results were confirmed by the following test: We divided
magnetic field maps in Fig. 15 in regular 11 times 11 square grid of
points. From field values in grid points we found the best fitting dipole
for both field components. Results displayed in Fig. 16 show that we obtained similar equivalent current dipoles in both cases. Like in the case
of single equivalent dipole fit in Fig. 13, they are approximately 6.7 cm
from the sphere center (2.3 cm from the sphere surface) and they are
positioned in the same region within ~1 cm and oriented in a similar
direction.
Discussion
In this paper we described two multichannel systems for detecting
MEG signals with atomic magnetometry. A unique aspect of these systems is that they use many common elements, so a multichannel magnetic field mapping can be realized with a relatively low complexity and
cost. The geometry of atomic magnetometers also offers new possibilities. In one of the geometries we measured the radial magnetic field as
a function of distance away from the source, which allows one to determine the magnetic multipole order or disentangle multiple dipole
sources. In another geometry we measured the two tangential components of the magnetic field using the same sensor. Unlike SQUIDs, atomic magnetometers do not suffer from cross-talk and can be used to map
vector magnetic fields in 3-D with millimeter resolution.
Atomic magnetometers also present unique challenges for MEG detection. Since the effective sensor positions are defined by laser beams,
they need to be determined in-situ. We apply known magnetic field
gradients to localize each of the channels based on detected magnetic
fields. The sensitivity and bandwidth of each channel is determined
for each data recording session by applying calibrated magnetic fields
at varying frequencies. The sensitivity of the sensors can reach 4 fT/
Hz1/2 above 10 Hz and magnetic field noise cancelation has been demonstrated between multiple channels.
We have successfully detected evoked auditory magnetic fields and
demonstrated source localization using the homogeneous conducting
4
80
3
60
40
20
1
0
ΔB, fT
Z, cm
2
0
−20
−40
−60
−1
−80
−2
−5
−4
−3
−2
−1
0
1
2
Y, cm
Fig. 11. Spatial correction of the distorted image on the photo detector array based on gradient imaging. Channels with good signal to noise ratios which are used in data analysis
are shown with large marks. The non-uniformity of the effective location of the channels
is due to optical distortions near the edge of the cell.
−100
−120
−0.2
0
0.2
0.4
0.6
t, s
Fig. 12. Auditory evoked magnetic gradients in By obtained with AM after averaging. The
auditory stimulus was at t = 0 s.
150
K. Kim et al. / NeuroImage 89 (2014) 143–151
(a)
(b)
Fig. 13. Magnetic field maps for the tangential, By (left panel), and radial, –Bx (right panel), components calculated from the best fitting equivalent current dipole. The current dipole has
been obtained by fitting measured signals at time t = 0.102 s after the stimuli from ten selected By gradiometer channels, defined as difference between the selected channels ◊ and the
reference (x). Δ is step between map lines, m and M denote maximum and minimum field values in fT, red, green and blue lines represent positive, zero and negative field values, respectively. Dotted line denotes the head model with radius 9 cm, as well as nasal and neck regions.
(a)
(b)
Fig. 14. The estimated MNE current distribution in two different views. The head model is a sphere with 9 cm radius, shown with orange open triangles. The reconstruction area of the
current distribution is shown with black triangles shaded with grey (spherical cap surface with height hc = 5 cm and radius rc = 8 cm). The estimated current distribution is shown
with cyan arrows.
sphere model. The localization of the equivalent current dipole corresponds well with the known position of the auditory evoked brain activity. While such model is relatively simplistic, it demonstrates that
atomic magnetometers have sufficient signal-to-noise ratio and spatial
mapping capabilities to be used for MEG studies. Our retro-reflected
(a)
probe beam scheme is expected to be very useful not only for the
MEG measurement but also for all other biomagnetic applications like
MCG since it can measure the tangential component of the magnetic
field for which the measured field is the maximum just above the
source, this is advantageous in source localization.
(b)
Fig. 15. Magnetic field maps calculated from the estimated current distribution shown in Fig. 14 for By (left panel) and –Bx (right panel).
K. Kim et al. / NeuroImage 89 (2014) 143–151
(a)
151
(b)
Fig. 16. Calculated magnetic field maps for the current dipoles obtained by fitting magnetic field maps shown in Fig. 15. By (left panel): RE = 0.21, CC = 0.982, rp = (−6.5,1.55,−1.17)
cm, |rp| = 6.8 cm, p = 55.2 nAm and ϕ = 25°, and –Bx (right panel): RE = 0.2, CC = 0.98, rp = (−6.25,2.19,−0.79) cm, |rp| = 6.7 cm, p = 54.3 nAm and ϕ = 29.7°.
Conclusion
The successful measurement and localization of human brain activity with a multichannel atomic magnetometer system demonstrates that
this non-cryogenic magnetometer is a possible alternative to SQUID
sensors. The current system is still relatively immature compared with
commercial SQUID systems but represents the first step toward the development of an economical and flexible multi-channel atomic MEG
system. We discussed specific procedures and technical challenges
that can guide future development of such systems.
Acknowledgment
The authors would like to acknowledge the financial support of NIH
and Packard Foundation (US), Korea Research Foundation and KRISS
(Korea), and Slovene Research Agency and MORS (Slovenia).
References
Cheyne, D., Bostan, A.C., Gaetz, W., Pang, E.W., 2007. Event-related beamforming: a robust
method for presurgical functional mapping using MEG. Clin. Neurophysiol. 118 (8),
1691–1704.
Cohen, D., 1972. Magnetoencephalography — detection of brains electrical activity with a
superconducting magnetometer. Science 175, 664–666.
Colon, A.J., Ossenblok, P., Nieuwenhuis, L., Stam, K.J., Boon, P., 2009. Üse of routine MEG in
the primary diagnostic process of epilepsy. J. Clin. Neurophysiol. 26 (5), 326–332.
Darvas, F., Pantazis, D., Kucukaltun-Yildirim, E., Leahy, R.M., 2003. Mapping human brain
function with MEG and EEG: methods and validation. NeuroImage 23 (Suppl. 1:S),
289–299.
Gusarov, A., Levron, D., Paperno, E., Shuker, R., Baranga, A.B., 2009. IEEE Trans. Magn. 45,
4478–4481.
Johnson, C., Schwindt, P.D.D., Weisend, M., 2010. Magnetoencephalography with a
two-color pump-probe, fiber-coupled atomic magnetometer. Appl. Phys. Lett.
97, 243703.
Johnson, C.N., Schwindt, P.D.D., Weisend, M., 2013. Multi-sensor magnetoencephalography with atomic magnetometers. Phys. Med. Biol. 58, 6065–6077.
Kakigi, R., Koyama, S., Hoshiyama, M., Kitamura, Y., Shimojo, M., Watanabe, S., 1995. Painrelated magnetic fields following painful CO2 laser stimulation in man. Neurosci. Lett.
192 (1), 45–48.
Kim, K., Lee, Y.H., Kwon, H., Kim, J.M., Kim, I.S., Park, Y.K., 2004. Optimal sensor distribution for
measuring the tangential field compoments in MCG. Neurol. Clin. Neurophysiol. 60, 1–4.
Kominis, I.K., Kornack, T.W., Allred, J.C., Romalis, M.V., 2003. A subfemtotesla multichannel atomic magnetometer. Nature 422, 596–599.
Kwon, H., Kuriki, S., Kim, J.M., Lee, Y.H., Kim, K., Nam, K., 2005. MEG study on neural activities associated with syntactic and semantic violations in spoken Korean sentences.
Neurosci. Res. 51 (4), 349–357.
Lee, Y.H., Kwon, H., Kim, J.M., Kang, C.S., Kim, K., Kim, I.S., Park, Y.K., Lee, S.G., 2003. A multichannel SQUID magnetometer system based on double relaxation oscillation
SQUIDs. IEEE Trans. Appl. Supercond. 13 (2), 755–758.
Li, Z., 2006. Development of a Parametrically Modulated Serf Magnetometer. Thesis
(Ph. D.) University of Wisconsin-Madison.
Okada, Y., Pratt, K., Atwood, C., Mascarenas, A., Reineman, R., Nurminen, J., Paulson, D.,
2006. BabySQUID: a mobile, high-resolution multichannel magnetoencephalography
system for neonatal brain. Rev. Sci. Instrum. 77, 024301.
Pantev, C., Hoke, M., Lütkenhöner, B., Lehnertz, K., 1987. Influence of stimulus intensity on
the location of the equivalent current dipole in the human auditory cortex. In:
Atsumi, K., Kotani, M., Ueno, S., Katila, T., Williamson, S.J. (Eds.), Biomagnetism, 87.
Tokyo Denki University Press, pp. 146–149.
Parker, R.L., 1977. Understanding inverse theory. Ann. Rev. Earth Planet. Sci. 5, 35–64.
Reeves, G., 2009. Smoothing Function for 2D Data ßmooth2a. Division of Biology Caltech
(March, http://www.mathworks.com/matlabcentral/fileexchange/23287-smooth2a).
Sander, T.H., Preusser, J., Mhaskar, R., Kitching, J., Trahms, L., Knappe, S., 2012. Magnetoencephalography with a chip-scale atomic magnetometer. Biomed. Opt. Express 3,
981–990.
Sarvas, J., 1987. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys. Med. Biol. 32, 11–22.
Seltzer, S.J., Romalis, M.V., 2004. Ünshielded three-axis vector operation of a spinexchange-relaxation-free atomic magnetometer. Appl. Phys. Lett. 85, 4804.
Shah, V., Romalis, M.V., 2009. Spin-exchange relaxation-free magnetometry using elliptically polarized light. Phys. Rev. A 80, 013416.
Shah, V., Vasilakis, G., Romalis, M.V., 2010. High bandwidth atomic magnetometery with
continuous quantum nondemolition measurements. Phys. Rev. Lett. 104, 013601.
Souza, M.J., 2008. A method for making large rectangular optical cells for SERF magnetometers. Fusion 55, 17–21.
Tarkiainen, A., Helenius, P., Salmelin, R., 2003. Category-specific occipitotemporal activation during face perception in dyslexic individuals: an MEG study. NeuroImage 19
(3), 1194–1204.
Xia, H., Baranga, B.A., Hoffman, D., Romalis, M.V., 2006. Magnetoencephalography with an
atomic magnetometer. Appl. Phys. Lett. 89, 211104.