LOCATOR



Locator 2

Starting up 2

Open Locator. 2

Open the appropriate sequence file 3

Place the neck collar with the three receivers around the subject's head 3

Place the transmitter near the subject 3

Sterilize the tip of the stylus 4

Begin the digitizing 4

Digitize electrode/marker coils 4

Digitize Head Shape 8

Finishing up 10

Troubleshooting 10

Hardware/software 10

Marker coils 11

Head shape 11

MRI from Locator Data 12

Open the Create MRI file dialog box 12

Load the Head Shape file 13

Load the Marker Coil file 14

Name and save the MRI file 14

Marker coils 15

Running the marker coils 15

Plug in and turn on the connection box. 15

Close and lock the MSR 15

Run the Marker Measurement 16

Locate the burst/coil waves 16

Locate the burst/coil waves 17

Sample Marker positions 18

Coil 0 18

Coil 1 18

Coil 2 18

Coil 3 18

Coil 4 18

Processing the Marker Coils (MEG160) 19

Open the Marker Positioning dialog box (if necessary). 19

Load the marker.sqd file (if necessary). 20

Make sure the estimation parameters are correct (if necessary) 20

Execute the localization function (if necessary) 20

Review the results of localization 21

Optional 22

Compare values with another marker.sqd file 22

Average two or more marker.sqd files 22

Locator

If localization of dipoles is to be performed, the localization programs (EMSE, Curry, Brainstorm, MEG-MUSIC) require information on the location of the sensors, the head, and the head with respect to the sensors (N.B., if EEG recordings are going to be used in the localization process, the positions of the electrodes too need to be digitized.). This information is pieced together by the markerwork process. The final product is a coregistered probe file that EMSE, Curry, and Brainstorm use to tell where the sensors and where the head were located. Information about the location of the head comes from the *marker.elp file that is produced by Locator during the head digitization process; this information is in a 3-D coordinate system defined by the head/anatomical markers, and varies from person to person. Information about the location of the MEG sensors is produced by the markerwork process and comes initially from the*marker.sqd file that was recorded by MEG160 prior to testing and ends up in the*meg.elp file produced by the markerwork process; this information is in a 3-D coordinate system defined by the center of sensor space. There are thus two different coordinate systems in play, one head-centered (*marker.elp) and one sensor-centered (*marker.sqd/*meg.elp); EMSE takes both of these files and creates a coregistered probe file that includes both head and sensor information in a single coordinate system, a head-centered frame of reference. It only needs three marker coil locations to perform this, and it disregards extra marker coils. Additionally, EMSE simply performs a rigid transformation between the coordinate frames. If there was a problem with any of the marker coils, either during digitization or marker.sqd recording, it can be disregarded as long as three coils worked.

The locator program is has been adapted to work with EMSE; however, both Curry and MEG160 can use the data it records if converted to the proper format. There are two outputs needed from the Locator program, head shape (.hsp files) and marker coil/electrode placement (.elp files). NB: If for any reason one of the marker coils or electrodes should become detached, either during digitization or during the experiment itself, technically the electrode/marker coil and head shape data should be re-digitized: if strict localization were going to be attempted, there would be a slight mismatch between where a coil or electrode was when first attached, and then later reattached. For the same reason, even if a subject has participated in many MEG studies, electrode/marker coil placement and head shape data should be collected anew each time.

Starting up

Open Locator.

On the desktop of the PC, there is a shortcut to "locator.exe;" use this to run the locator. The actual path of the program is C:/source signal imaging/Locator41. Sometimes, the computer will not be able to find the Fastrak; if this occurs, a message will come up. If this happens, turn off the Fastrak (the switch is in the back), wait 10 seconds, then turn it on again. When the green light on the front of the Fastrak becomes steady (stops blinking), try running Locator again.

Open the appropriate sequence file

The sequence file identifies the anatomical landmarks, marker coils, and/or electrodes for which data will be recorded. Under the "File" menu, select "Open" and choose the appropriate sequence file. There are at present 3 sequence files:

mark-5coils.seq" which is for the 5 electromagnets and fiducials

"mark-EEGonly.sqd" which is for the 23 electrodes of the E-cap and fiducials,

and "mark-5coils-EEG.seq" which is for the 5 electromagnets, 23 electrodes of the E-cap, and fiducials.

Place the neck collar with the three receivers around the subject's head

This is easiest if you ask the subject to hold the marker coil wires that have been attached to his/her head, and you slide the collar underneath. Make sure the subject's chin goes in the chin rest; otherwise, the subject will be very uncomfortable. Velcro the collar shut.

Place the transmitter near the subject

Make sure the transmitter at the right rear corner of the chair (as you face the chair). On the top of the transmitter are two axes in one corner, one designated "X," the other "Y;" the subject's head should be within the box traced by these axes, and these axes should be aligned as directed on the floor. If the subject's head is not within a box formed by these axes, the locator program will not be able to run as efficiently. Note also that these X and Y coordinates are totally unrelated to the XYZ coordinates that are used in the EEG and MEG systems.

Sterilize the tip of the stylus

Use an alcohol swab; there are extra swabs located in the top left-hand drawer of the desk in E10-039.

Begin the digitizing

Select, from the "Digitize" menu the "number of receivers on subject" submenu, and select 3 (this refers to the three located on the neck collar).

Select, from the "Digitize" menu, the "electrode data" option. The locator program will now walk you through electrode/marker coil digitization.:

First, it asks that you confirm that the subject is wearing the collar, and that the transmitter has been placed near the subject.

[pic]

[pic]

Digitize electrode/marker coils

Electrode/marker coil files (.elp) record the locations of three anatomical landmarks (fiducials) (the nasion/NZ, left preauricular point/LPA, and right preauricular point/RPA) that are used to match up structural MRI data to MEG data, the five marker coils (red-#0, yellow-#1, blue-#2, white-#3, ,green-#4) that are used to match up MEG sensor placement with respect to the subject's head, and/or the 23 electrodes in the E-cap. Note: the blue marker coil (#2) is typically placed above the NZ (the NZ is the intersection of the vertical crease between the eyebrows and the horizontal start of the nose cartilage), the red marker coil (#0) is typically on the LPA (the LPA is on the subject's left; it can be found by asking the subject to move his/her jaw up and down and left and right. The point where the jawbone goes into the skull, right by the ear, is the LPA), and the yellow marker coil (#1) is typically placed on the RPA (the RPA is identical to the LPA, only on the subject's right).

• After selecting the "electrode data" option from the "Digitize" menu (see above), the locator program (after confirming set up is appropriate) will ask for the NZ. Place the tip of the stylus on the nasion (NZ) and press the stylus button once.

• The locator will now prompt for the LPA; place the stylus on the red marker coil (LPA) and press the stylus button.

• The locator will now prompt for the RPA; place the stylus on the yellow marker coil (RPA) and press the stylus button

• The locator will now proceed into the appropriate sensor digitization(s) (that is, marker coil, electrode, or both, depending on which sequence file was selected—see above) digitization.

• Marker Coils: The order is as follows: mark0 is the red coil (LPA); mark1 is the yellow coil (RPA); mark2 is the blue coil (above Nasion); mark3 is the white coil, on the subject's left forehead; mark4 is the black coil, on the subject's right forehead

• Electrodes/E-cap (see diagram below): The order is as follows—down the left hemisphere (FP1, F7, F3, T3, C3, T5, P3, O1), to the midline (FZ, Ground, CZ, PZ), down the right hemisphere (FP2, F8, F4, T4, C4, T6, P4, O2), to reference/auditory (A1, A2), to the ocular sites (EOG1, EOG2)

Electrode placement diagram

When you have finished, you should see a figure as below, showing the 3 fiducials and five markers (the labels for the markers may be turned off or on using the Labels radio buttons in the lower right corner.

[pic]

• When you have finished digitizing the fiducials and electrodes and/or markers, the locator will ask you if you will to proceed to head shape digitization. When done with electrode data, the program will automatically ask you if you wish to proceed with head shape digitization; if head shape is desired, press “yes.”

Digitize Head Shape

Head shape files (.hsp) record the approximate shape of a subject's head by marking a point every .5cm the stylus is moved (this is modifiable—the distance between points can be increased or decreased). Later, during localization, these points are used to set limits: Localization algorithms are typically “dumb” in the sense that they do not care whether the source they find is inside the brain or in space; so, the head shape data is used to constrain localization algorithms to preclude/exclude localizations outside the brain. Some localization algorithms compute from the head shape data the best-fitting sphere, and then constrain localizations to be within that sphere.

Ideally, data points should be collected from all over the target area, not just in one location—that is, the whole of the skull enclosing the brain, not just the top of the head. When collecting head shape data for MEG, the target area is the upper portion of the head (i.e., the portion housing the brain; additionally, the nose may be digitized to serve as a landmark. The more data points collected, ostensibly the more accurate the localization constraint computed; currently, we try to get around 600 data points, although in the past anywhere between 500 and 1000 points have been collected. The number of points collected is displayed in the lower right corner of the window.

Head shape data is easier to digitize in a number of small sequences of points, rather than a single sequence of 500 or more points. In order to digitize a head shape sequence, the stylus should be placed so that its tip is touching the head. The button on the side of the stylus should be pressed once to begin recording points. After a predetermined distance has been traveled, another point is located; as long as the stylus is moving, points will be recorded. If a zigzag path is traveled (by moving the stylus in short jags left and right), the number of points recorded will be maximized. To finish digitizing a sequence, press the button on the stylus a second time.

If you accidentally remove the stylus from the head while digitizing, or if a sequence looks odd, you may delete it by pressing the delete sequence button in the lower right. Sometimes, the digitizer will itself recognize that a series of points is unusual, and ask you if you wish to remove them.

When head shape digitization has been completed, the "accept" button (bottom middle of the window) should be pressed. A completed head shape should give a reasonable approximation of the upper portion of the head.: [pic]

Finishing up

After digitizing electrode/marker coil and head shape data, two files should be saved manually via the "File" menu. Save the head shape data by selecting an ".hsp" file type (bottom of the save window). Save the electrode/marker coil data by selecting an ".elp" file type (bottom of the save window). When exiting the locator program, it will ask again to save the data, due to a flaw in the program; it gives this prompt whether or not the data files have already been manually saved (relax, if you have in fact manually saved the .hsp and .elp files, you can ignore this message, however stressful that may be. If you are feeling very insecure, you can look in the data folder (C:/fleissbackup/locator41/data/) to make sure saved the marker.elp file was saved).

Remove the collar from the subject, being careful not to accidentally rip off one of the marker coils (as with putting the collar on, this works easily if the subject is asked to hold up the wires leading from the marker coils).

The PC and the digitization hardware may be left turned on.

Troubleshooting

Hardware/software

• Monitor problems: Occasionally, the monitor of this PC freezes or is frozen; if this occurs, simply turn the monitor off and on again

• PC problems: Occasionally the computer itself will freeze; in this case simply restart the computer (note: if you restart the computer, you will have to log on again. There is no password to log on; when logging on, user the username and leave the "password" section blank); normally, when the PC is rebooted, Locator requires that the Polhemus 3space digitizer be turned off and on again

• Locator program problems: Occasionally also, the locator program will tell you that it cannot "find" the digitization hardware; if this happens, try turning the digitization hardware off and on, and also make sure all the connections on the back of the digitization hardware and the PC are firm-- it is unclear why, but these two steps normally fix the problem. If both of these steps fail, try restarting the PC (see above note about logging on)

Marker coils

• Pressing the stylus button too soon—it is easiest to quit the program and begin again; while it is possible to edit the .elp file, it is much more time-consuming than simply starting over.

• Digitizing coils in the wrong order—see previous

• Subject movement-- Do not worry if the subject moves: the three sensors in the collar are designed to compensate for small movements. If you make a mistake digitizing the electrode placement, it is best to quit the program and begin again.

Head shape

• Stylus lifted from the head—if sufficiently deviant, the Locator program will offer a warning and ask if the sequence should be deleted. Even if the Locator program does not offer the option of deleting the sequence, in the lower right corner of the screen is an option for manually deleting the sequence. NB: the first sequence of points digitized cannot be deleted; so, if the first sequence of points is inaccurate, it is best to "cancel" (bottom middle of the window) out of the digitization process and begin anew.

• Subject movement-- As with marker coil digitization, do not worry if the subject moves: the three sensors in the collar are designed to compensate for movement.

MRI from Locator Data

To create an "mri" file from Locator data, you need the marker.elp file, the shape.hsp file, and MEG160.

Open the Create MRI file dialog box

Select Create MRI File from digitizer data either from the File menu or the MRI menu

[pic]

Load the Head Shape file

[pic]

Load the Marker Coil file

[pic]

Name and save the MRI file

[pic]

Marker coils

Running the marker coils

Plug in and turn on the connection box.

The connection box is where the marker coils are plugged in. It also provides power to the electromagnets (via 8 AA batteries), and feeds in the DAQ as well (channel 104). Turn on the connection box inside the MSR before beginning your recoding session. Plug in the marker coils according to the color coding.

Close and lock the MSR

Run the Marker Measurement

This can be done simply by pressing "start" under the Marker Measurement section of the Monitors and Acquisition dialog. After 16 successful epochs have been recorded, the process will terminate. Consult the "MEG/How to use it" manual for information on the Monitors and Acquisition dialog box. To view the process as it occurs, see below.

Locate the burst/coil waves

You should see the "burst" waves, five in a row, repeated. Each wave represents one of the markers. Sometimes they will be smaller, depending on where the subject's head is. If you are not sure that you see all five, examine channel 104, the marker coils reference channel. If you do not see 5 waves there, the coil set is not working properly.

[pic]

[pic]

Sample Marker positions

Coil 0

[pic]

Coil 1

[pic]

Coil 2

[pic]

Coil 3

[pic]

Coil 4

[pic]

Processing the Marker Coils (MEG160)

After running and recording the marker coils, you should process the file. MEG160 will prompt you for automatic processing. Otherwise, use the manual method to invoke the Marker Positioning below. This will show where the coils are located, and thus give you an idea of where the subject's head is located.

Open the Marker Positioning dialog box (if necessary).

Open the Marker Positioning dialog box, either via the File menu, or the MRI menu

Load the marker.sqd file (if necessary).

[pic]

Make sure the estimation parameters are correct (if necessary)

Number of Markers should match the number of working coils.

Reference Channel is 104 for the 96-channel system, and 72 for the 64-channel system.

Averaging Count should be 16.

Trigger Threshold should be around 160.

Execute the localization function (if necessary)

Review the results of localization

In particular, examine the GOF (goodness of fit) of the estimations (these should ideally be above 99%), and the locations of the coils as displayed in the Axial (top-bottom), Sagittal (side), and Coronal (front-back) views.

[pic]

Optional

Compare values with another marker.sqd file

If you have more than one marker.sqd file for the same session, you may compare values. Press the Compare button. You will be asked to load the second file. The differences between the localizations will be presented numerically:

[pic]

and may also be presented graphically if you answer "yes" to display:

Circles represent the current file, and crosses represent the second file.

Average two or more marker.sqd files

If you have more than one marker.sqd file from a session, after estimating each file, you may create a Grand Average (consult the MEG160 manual for information on this) of the estimated files. This grand average file will also average the estimated placement of the marker coils.

-----------------------

LPA (left preauricular point)

RPA (right preauricular point)

NZ (nasion)

EOG2 (right eye)

EOG1 (left eye)

A1

O2

T6

A2

T5

O1

P3

P4

CZ

PZ

F4

GND

C4

C3

T4

T3

F3

FZ

F8

F7

FP2

FP1

LPA/marker 0

RPA/marker 1

marker 4

marker 3

marker 2

Power switch: turn on before use; TURN OFF AFTER USE

5 magnets, on subject (note: NZ/coil2 is weaker than others, but it is still strong enough to localize)

Coil 0 (LPA)

Coil 1 (RPA)

Coil 2 (NZ)

Coil 3

Coil 4

Here, coil 3 (white/left forehead) is not working; not only is it not seen under the sensors (upper picture), it is also not present in the feeback wave on channel 104 (lower picture)

Coil 2 (aboveNZ)

Coil 0 (LPA)

Coil 1 (RPA)

Coil 4

Coil 3

[pic]

x-axis

y-axis

[pic]

[pic]

Press the ">>" button to advance

[pic]

[pic]

[pic]

[pic]

[pic]

New points

Old points

Total points

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

[pic]

This data will appear in the Heasd Shape Summary

[pic]

This preview will appear

[pic]

[pic]

[pic]

Nazion

+

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download