top bar

Difference between revisions of "MicroMagellan"

(Drift Compensation)
(Setting Up Multiple Acquisitions)
Line 155: Line 155:
"Run all" will sequentially complete all acquisitions listed.  
"Run all" will sequentially complete all acquisitions listed.  
If you would would like to collect time data in different areas concurrently (rather than completing a time lapse in one area, and then beginning the second time lapse in another area), select “interleave” and then  “Run all”
If you would would like to collect time lapse data in different areas in parallel (rather than completing a time lapse in one area, and then beginning the second time lapse in another area), select “In parallel” and then  “Run all”

Revision as of 13:48, 8 June 2016

The Micro-Magellan plugin allows flexible, dynamic device control and acquisition. Explor acquisitions enable navigation of samples in XY and Z, to build a 3D map of sample quickly and with minimal photobleaching. Surfaces allows you to define an non-cuboidal region of interest and define sample morphology, which can then be used to define sample-driven parameters such as laser power at distance from surface. Using grids or surfaces, you can designate multiple areas to image in a single acquisition in series or in parallel. A high level overview of the capabilities of Micro-Magellan can be found here:

Getting Started with Micro-Magellan

In order for Micro-Magellan to operate, it needs to know the direction of Z movement. This is defined in the Hardware Configuration Wizard.


In order for Micro-Magellan to acquire tiled images and assemble them, it must have an accurate pixel size calibration for the objective in use. If you have already calibrated your pixel size using the “Pixel size calibrator” plug-in from MicroManager, this should automatically propagate to Micro-Magellan. Before using Explore mode, verify that the calibration is correct by pressing “Calibrate”. If the fields are populated as below, Micro-Magellan has the necessary information. These parameters can be adjusted manually to fine tune stage movement for improved stitching. If the calibration window shows a size of 0.0um, you can run click Start and follow the wizard instructions to generate an affine transform.


Setting Up Micro-Magellan for Device Control

From the plug-ins menu, select Micro-Magellan, and the unpopulated window opens. Select "Configure Device Control"


A new window will open with a list of available devices and properties. You may also enter a "nickname" that more easily identifies the device/property. Selected devices/properties will appear in the Device status/control tab after pressing "Ok"


Explore Acquisitions

Before exploring your sample you will need to enter your desired Z-step size and tile overlap. The channel group drop down is analogous to the channel group drop down within MicroManager’s Multi-Dimensional Acquisition and will be pre-populated with the configurations groups you have created. Select the settings group that contains your channel configurations.


Pressing “Explore!” will open a new window: ExploreWindow.jpg

On the left side is your image area; a field of view will highlight blue as you hover over it. On the right, are look-up tables (LUTs) for each channel and several other acquisition setup parameters. To acquire an image in that area click to select (area will highlight green during active selection), once your selection is highlighted magenta, click again to confirm. You can click and drag to select multiple fields of view to be imaged. The "Z limits" scroll bars below the image window control which Z locations will be imaged, and will adjust automatically to the selected Z-step size.


Once you have begun to image, the Z-plane currently being displayed will appear dark green in the Z-limits scroll bars. Other acquired Z-planes will appear light green. In the example above, a Z-stack has already been collected (light green), and additional Z-planes will be collected between the new Z-limits.

You can review acquired images by scrolling with the “Z” bar, and right-clicking and dragging to navigate in the XY plane. You can also zoom in and out by using the mouse scroll wheel with the image window selected.

Once you have explored your sample and would like to select areas to image over time, with more optimized parameters, better Z-resolution, etc. you can use either the “Grids” or “Surfaces” tab above the LUTs to do so.


In the Grids tab, you can define an N x N grid, which will appear over your explored image. With the Grids tab active, you can drag this grid over the area you would like to image. You may generate multiple grids of different sizes by pressing “New Grid”, and then move them to different areas.



All of the grids generated in explore mode will populate the “grids” tab at the top of the Micro-Magellan main window, next to the Device status/control tab. If you are creating multiple grids, it may be helpful to toggle back to the Micro-Magellan main window, and select and rename the grids with their identifying features (ie: airway, duct, tumor).


The surfaces tab allows you to create and edit surfaces, which can be used to define non-rectangular acquisition volumes or automatically vary hardware settings based on sample morphology during acquisition. Surfaces are created by marking interpolation points on 2D slices of an image. Micro-Magellan then uses these points to interpolate a continuous surface.

A new surface will be created automatically after clicking on the surfaces tab. A new surface can also be created by clicking on the "New Surface" button. Left clicking on the image screen with an active surface adds points. Left clicking on a point will remove it.


The Z-slider is then moved down further into the sample to specify an additional 2D slice of interpolation points.



All of the grids and surfaces generated in Explore mode will populate the "grids" and "surfaces" tabs at the top of the Micro-Magellan main window, next to the Device status/control tab.


Below the Explore mode settings, there are a series of tabs to manage acquisition setup.


The "Save" tab allows you to define the name of the acquisition and uses the same directory as the explore mode. To set up a simple single area acquisition, enter a name fore the file.


In the "Time" tab, you can enter a time interval between image acquisition and the total number of time points to acquire



In the “Space” tab there are several options for defining the volume you want to image. In a simple Z-stack you enter start and end Z positions, and either the grid or surface to define XY location and positions. If you have created surfaces, you can also elect to image the area between two defined surfaces with a buffer on either side.


XY footprints can be either grids or surfaces, and define the XY positions that are used for a given acquisition. The XY positions can be changed during acquisition by swapping or modifying XY footprints. However, all XY positions imaged at any point during the acquisition must be present in its footprint at the start. Positions can be removed by changing the footprint and later re-added, but new positions cannot be created during acquisition to expand the imaged area beyond its original XY bounds. Practically, this means if imaging a sample that is expected to move or change over the acquisition, it is a good idea to use an XY footprint containing all possible XY positions where the sample might be imaged. If only a subset of the positions need to be imaged at the initial time point, the acquisition can be immediately paused after starting, and the XY footprint adjusted to a smaller area.


The Z slices that are collected for a given acquisition can be defined in several ways:

Simple Z Stack

The same Z-slices are imaged at every XY position, creating a tiled cuboidal volume

Volume between two surfaces

Supply two surfaces to bound the top and bottom slices to collect at each Z stack. This is useful for defining 3D imaging volumes with complex shapes.

Within distance from Surface

Z slices are collected within a given distance above a below a surface. This is useful, for example, for collection a given distance from the top of a tissue, or for imaging a rectangular region that is tilted relative to the optical axis.


The “Channels” tab allows you select which of the available channels to collect, and the exposure time for each.

Covaried Settings

Covaried settings expose all properties of hardware devices and allow users to automatically vary these settings prior to the acquisition of each image based on the state/value of each other. In addition, calculations based on the geometry of interpolated surfaces can also be included in a covariant pairing, enabling automated variation of hardware settings based on surface morphology.

A covariant pairing consists of an independent covariant, a dependent covariant, and a list of pairs of values between the two. Before the acquisition of each image, the value of the independent covariant is either read from the hardware or calculated from the the associated surface. The value of the dependent covariant is then calculated and set based on the independent-dependent pairings supplied in the setup window. If both the independent and dependent are numerical, then the value calculated for the dependent covariant comes from a linear interpolation of the supplied pairings. If either the independent or dependent are discrete, a value is only set if the current value of the independent covariant is also present in the supplied pairings.

The “+” button on the left brings up the window below, allowing you to select one hardware setting as the independent variable, and a second as the dependent variable. 


In the example below, the position of the Z-stage determines the EOM voltage (laser power), so that laser power can be increased to compensate for scattering deeper in the sample.


To set interpolation points for the covaried settings, select the "+" button at the bottom right. If you have defined surfaces, you can also choose to vary hardware settings based on surface data.


For a given tissue type and morphology, these parameters may be optimized and saved for use across samples. To load a saved set of interpolation points and settings, click “load” and select the saved parameter file. You will be prompted to select which surface you would like to assign the settings to, and which surface parameter you would like to use.


Drift Compensation

In 3D imaging experiments, stable imaging is often beset by focus drift as a result of morphological changes in the sample or thermally induced focus drift. Micro-Magellan’s drift compensation module provides automated tracking of sample movement in the Z axis based on the 3D cross correlation between images from a designated fiduciary channel over the previous 2 time points.

This feature can be set up on the “Drift Compensation” panel in the “Acquisition Settings” tab of the main Magellan GUI. The use channel dropdown allows users to select which channel will be used as a fiducial for calculation of cross correlation. Second harmonic generation signals in multiphoton imaging are often a useful choice this signal due to their photostability, but any other stationary, photostable structure should be usable provided that there is not significant bleed through from mobile signals in other channels.

-The “maximum displacement” field is a safety feature that sets an upper limit on the cumulative drift correction that occurs over an acquisition, to prevent unexpected changes in a fiducial signal from damaging objective lenses. -The “Drift compensation z device” allows the user to select which focus drive will be used to offset drift. At present, this must be a different focus drive from the primary one used in acquisition. -The “Set initial position” box is useful when running multiple acquisitions in parallel that each use drift compensation. Since the position of the sample in the coordinate space of the primary focus drive will change based on the position of the drift compensation focus drive, it is sometimes to provide different initial positions for the drift compensation drive. For example, this feature would allow the positioning of the full range a piezo z drive at different absolute regions of the sample Z coordinate space, and each of these to be drift corrected during time-lapse acquisition independently.


Setting Up Multiple Acquisitions

To acquire more than a single grid or surface at a time, select the “Setup multiple acquisitions” tab. Click “+”, this will add a line with the name currently entered in the “Saving” tab below. Enter the parameters for this acquisition in the other acquisition settings tabs.


Clicking “+” again will add another line, it will appear with the same name. Highlight the second (new) acquisition and change its name in the “Saving” tab below, and change necessary parameters in the other acquisition settings tabs. "Run all" will sequentially complete all acquisitions listed.

If you would would like to collect time lapse data in different areas in parallel (rather than completing a time lapse in one area, and then beginning the second time lapse in another area), select “In parallel” and then “Run all”


Using Micro-Magellan with BigDataViewer

Micro-Magellan writes data in a multiresolution file format that is compatible with FIJI's BigDataViewer (, a reslicing browser for terabyte-sized image volumes that is integrated with a number of image processing tools. In order for BigDataViewer to recognize the Micro-Magellan file format, the "Magellan.jar" file must be copied into a FIJI installation. To do this:

1) Open FIJI and select Help--Update
2) Click the "Manage Update Sites" button
3) Check the box next to Micro-Magellan
4) Click "Close"
5) Click "Apply changes"

Open a Magellan dataset by selecting "Plugins--BigDataViewer--Open XML/HDF5" and select the "FIJI_BigDataViewer_Metadata.xml" file within a Magellan dataset

Micro-Magellan source code

The source code can be accessed in the Micro-Manager subversion repository at

User Guide:Media:Micro-MagellanUserGuide.pdf
Authors:Henry Pinkard, Kaitlin Corbin
Available since version:1.4.22
© Micro-Manager : Vale Lab, UCSF 2006-2011 | All Rights Reserved | Contact