Tiling large samples using Zeiss MultiTime Macro

[Gareth Howell]

Dear all

I have two questions relating to collecting tiles images on a Zeiss 510:

1) I have a user who wants to scan large areas of tissue using our confocal.
They want images at 1024 for good resolution. At the moment we are using the
Zeiss Stage Control tiling option but are limited to 4Kx4K pixel arrays. I
was wondering if someone had a protocol for using the MultiTime Macro within
LSM510 software to collect multiple panels of 4Kx4K pixels with the aim of
stiching them all together for one big image.

2) I'm probably missing a trick here but why can't I collect tiled images
using the camera within the LSM510 software; would it be possible with
AxioVision on the same microcope?

Regards

Gareth

[Daniel James White]

Hi Gareth,

Begin forwarded message:

> Date:    Fri, 19 Feb 2010 05:32:29 -0600
> From:    Gareth Howell <[hidden email]>
> Subject: Tiling large samples using Zeiss MultiTime Macro
>
> Dear all
>
> I have two questions relating to collecting tiles images on a Zeiss  
> 510:
>
> 1) I have a user who wants to scan large areas of tissue using our  
> confoc=
> al.
> They want images at 1024 for good resolution.

you should click the "optimal" button to get the proper number of  
pixels for the zoom and objective lens you are using.
then you get the best resolution the lends can see without missing  
info or having far toom many , too small pixels.

using 1024x1024 makes no sense in most cases.

> At the moment we are using =
> the
> Zeiss Stage Control tiling option but are limited to 4Kx4K pixel  
> arrays. =
> I

are you ....? I wasnt aware of that limitation?
cant you do 10x 10 tiles with the tile scan tool on the stage control  
window?

> was wondering if someone had a protocol for using the MultiTime  
> Macro wit=
> hin
> LSM510 software to collect multiple panels of 4Kx4K pixels with the  
> aim o=
> f
> stiching them all together for one big image.

the multi time macro can do that.... just read theinstructiosn and  
then good luck.
the latest version costs money from Zeiss,
but i have to say its not very easy to use and its a bit crashy...
i hoper its better with the Zen software on newer systems,

it works for us, but its not very pretty to use.

Actually for the actual stitching of the tile scans,
the imageJ - Fiji plugin by Stephan Preibisch is more sophisticated  
and works great
You cant see the join! and it does multi colour z stack tile stitching.
see here
http://pacific.mpi-cbg.de/wiki/index.php/Stitching_2D/3D


>
> 2) I'm probably missing a trick here but why can't I collect tiled  
> images=
>
> using the camera within the LSM510 software; would it be possible with
> AxioVision on the same microcope?

LSM 510 is a laser point scanning confocal microscope,
and has no camera.
It has photomultiplier tubes instead.

if this is not clear to you , then you need some help to learn about  
the difference between a widefield microscope and  a confocal  
microscope.
Else you may be wasting you time.....

see here for some help with that
https://ifn.mpi-cbg.de/wiki/ifn/images/6/62/Confocal_notes.pdf
and then
https://ifn.mpi-cbg.de/wiki/ifn/images/e/e9/Optical_Sectioning_ProsAndCons10-2009.pdf


cheers

Dan
Dr. Daniel James White BSc. (Hons.) PhD
Senior Microscopist / Image Processing and Analysis
Light Microscopy Facility
Max Planck Institute of Molecular Cell Biology and Genetics
Pfotenhauerstrasse 108
01307 DRESDEN
Germany

[phone] (German Mobile)
+49  (0)351 210 2627 (Work phone at MPI-CBG)
+49  (0)351 210 1078 (Fax MPI-CBG LMF)

http://www.bioimagexd.net                BioImageXD
http://pacific.mpi-cbg.de                Fiji (is just ImageJ - batteries included)
http://www.chalkie.org.uk
[hidden email]
( [hidden email] )

[Julio Vazquez]

Hi Gareth, 

Below is a summary of the procedures for collecting tiled images with Zeiss LSM. I wrote this a long time ago, so there may be some errors, or differences with different versions of LSM software, but it could be a start.

 Also, as Dan mentioned, LSM software acquires images by scanning a spot of light and reading intensities with a PMT, not a camera. However, if you have an extra camera port on your LSM, you could in fact add a camera and hook it up to a PC running Axiovision (or any other acquisition software). Most acquisition software nowadays have capabilities to collect large images through tiling and/or stitching, assuming you have a motorized stage. These images would not be confocal though....

===

Zeiss LSM 510 META NLO User's Manual. Part III.

Advanced Techniques for multiple location imaging.

Contents:

I. Collecting a tile image.

II. Calibrating stage.

III. Marking Multiple locations.

IV. Collecting Tiled Stacks. 

V. Collecting a grid of stacks.

VI. Collecting stacks at multiple locations.

VII. Collecting a Time series at multiple locations.

VIII. More complex imaging patterns.

I. Collecting a tile image.

1. Set up your sample and imaging parameters. Optimize image.

2. In the main menu, open the Stage control panel:  ACQUIRE>STAGE

3. At the bottom, under Tile Scan, enter the tile dimensions (n(x), n(y)). Maximum tile size is 4096x4096 pixels. Therefore, you can collect at most a 4x4 tile at 1024x1024 pixels, but 8x8 tile at 512x512. The lower the resolution, the larger the field of view you can collect.  

4. Save your tile image.

Notes: 

It is best to collect a small tile first at good resolution (e.g. 2x2 at 1024x1024 at zoom 1-2) to test stage alignment. If there is an offset in the tiles, proceed with stage calibration as described below.

This method only allows the collection in a single plane (not z stacks). For tiled z stacks, see section IV.

II. Calibrating stage.

You need a sample that has good contrast. 

1. Place sample on stage. Choose one channel that shows good image definition and brightness (overall good contrast). 

2. Select objective and zoom factor. Optimize image.

3. In main menu, select ACQUIRE>STAGE, and set for tile collection under "Tile Scan", e.g. 2x2 tiles.

Collect an image and check alignment in x and y. If Alignment is good, the stage is calibrated: proceed with your experiment. Otherwise:

4. In the main menu, select MACROS>Rotation

5. In the calibration menu window, click "Calibrate". Microscope will start automatic scans and attempt to calibrate stage. When calibration is done, a correction factor will be displayed.

6. Collect a new tile to test the new calibration. Repeat if you are not happy with result. You may also try to change the numbers manually until satisfied. Note that alignment may not be perfect.  

Note: Calibration will be best if you select "speed 1" in the Stage Position panel (Stage Control window).

III. Marking Multiple locations.

The Zeiss 510 allows you to mark locations in your sample and store them. Once stored, you can navigate to any of those locations for further imaging. 

The recording of stage positions is done within the Stage control window.  You can mark positions in two different manners: manually, or using a tile scan.

a. Manual marking:

1. You can select any position, for instance the center of your sample, and mark that as the "center": Find the position you want to be the center by visual inspection (e.g. you can use the 2.5x objective to do a rapid survey of the sample.

2. In the stage window, under "Stage Position", click "Zero", and click "Mark Pos."

3. Find a new location by visual inspection through the eyepiece,  or by moving the stage in fast scan mode. Find a new interesting location. Click "Mark Pos.". 

4. Repeat for all locations. 

5. The list of locations will be recorded in the stage control window, in the  "Marks" pull-down menu. 

6. You can now select a location in the pull-down menu, and use "Move To" to move the stage to that position.

Note: 

The Stage control window also remembers the z position. Therefore, you can mark different locations in x,y,z. I noticed, however, that if you mark one position in x,y,z, the system will not record a new position that has the same x,y coordinates, but a different z location.  Therefore, if you want to record two z positions at the same location, you will need to slightly move the stage in x or y. 

b. Tile marking.

1. Collect a tile.

2. In the Tile Scan panel, click "Mark"

3. Click on your tile scan image to mark different locations.

Collection of complex images.

The Multi-Time Macro allows the collection of single z sections, or z stacks, in a tiled manner, or on a grid pattern, or at multiple locations. The various procedures are described below.

IV. Collecting Tiled Stacks. 

1. Select a Multi-track configuration (not single track, as single track do not remember certain settings).

2. Set your sample and optimize image collection.  Adjust all parameters.

3. If you want z stacks, set the z stack parameters.

4. Save all the parameters as a new configuration: Click "Config", enter a new name (e.g. aa1), and "Store".

5. Calibrate stage as described in section II.

6. Start Multi-Time Macro  (Main Menu>MACROS>MultiTime"R")

7. Click "Refresh" button.

8. Load Multi-Track configuration from pull-down menu in Multi-Time window (Config, Time-Interval and Bleach), (e.g. aa1)

9. For a single tile, make sure "number of scans" is set to "0", and z-stack is checked.

10. Under "List of Blocks", set number of repetitions to "0"

11. Make sure autofocus is NOT checked, and bleach is NOT checked.

12. Click "Single location" button on top of Multi-Time window.

13. Click "Edit Locations"

14. Select "Tile"

15. Enter the tile size (e.g. 2x2)

16. Select Overlap (e.g. 20%). The overlap is required for automatic stitching of images. If Stitching fails, use higher overlap.

17. Click "Create Tile"

18. Click "Options" (in Multi_Time window); Check "Keep Final Image Open" and "Save Final Image"

19. Click "Image Database", and select the database where your final image will be stored.

20. Click "Start Time"

21. Wait until images are collected and assembled. The final image will be saved in the database you selected. Intermediate images will be saved in a temporary directory in drive C. 

V. Collecting a grid of stacks.

The "grid" function collects single sections or stacks in a grid pattern, with current position used as the origin.

To collect a grid, follow the same procedure as for a tile, except for:

14. Select "Grid"

15. Enter the spacing of the grid in x and y.

16. Click "Create Grid"

Proceed as for a tile with steps 18-21.

--

Julio Vazquez

Fred Hutchinson Cancer Research Center

Seattle, WA 98109-1024

http://www.fhcrc.org/ 

On Feb 21, 2010, at 9:16 AM, Dan White wrote:

LSM 510 is a laser point scanning confocal microscope,

and has no camera.

It has photomultiplier tubes instead.

[Lingqing Zhang]

Hi Gareth,

As long as you have a motorized stage from Zeiss you should be able to collect tiled images as any array combination as you want, and the stage move]ing range is the limit. But you have to use a third party software to stitch them together. I wrote a Matlab code to do so. I had done that two years ago at St. Jude Children's Research Hospital, and now forget the exact steps, but it was quite simple. The Zeiss software we used was LSM4.3.

Lingqing
**********************************
Lingqing Zhang, PhD
Nonlinear Optical Imaging Lab
LM/EM Imaging Core Facility
West Virginia University
1 Medical Center Drive
Morgantown, WV 26506
Phone: [phone]
Email: [hidden email]

On Fri, Feb 19, 2010 at 6:32 AM, Gareth Howell <[hidden email]> wrote:

Dear all

I have two questions relating to collecting tiles images on a Zeiss 510:

1) I have a user who wants to scan large areas of tissue using our confocal.
They want images at 1024 for good resolution. At the moment we are using the
Zeiss Stage Control tiling option but are limited to 4Kx4K pixel arrays. I
was wondering if someone had a protocol for using the MultiTime Macro within
LSM510 software to collect multiple panels of 4Kx4K pixels with the aim of
stiching them all together for one big image.

2) I'm probably missing a trick here but why can't I collect tiled images
using the camera within the LSM510 software; would it be possible with
AxioVision on the same microcope?

Regards

Gareth

[Ian Dobbie]

Lingqing Zhang <[hidden email]> writes:

> Hi Gareth,
>
> As long as you have a motorized stage from Zeiss you should be able to collect
> tiled images as any array combination as you want, and the stage move]ing range
> is the limit. But you have to use a third party software to stitch them
> together. I wrote a Matlab code to do so. I had done that two years ago at St.
> Jude Children's Research Hospital, and now forget the exact steps, but it was
> quite simple. The Zeiss software we used was LSM4.3.

ImageJ is also able to do simple tile stitching. Unfortunately it
expects the images to be ordered right-to-left top-to-bottom. When
collecting the data the stage move is often the major time impediment so
it is much more efficient to collect data left to right on the first
line, right to left on the second etc...

There are also a couple of packages that will do some kind of
correlation based image matching at the boarders to get the images
stitched in the right places (I think turboreg will do
this). Unfortunately my application was sparse cells at low res in
fluorescence so all there was on most boarders was noise!

I have hacked around the ImageJ stitching macro to do this (and also do
time series). If anyone is interested I can send them a copy of my
working, but not great code.

Ian

[Keith Morris]

Hi Gareth,

Well obviously ask Zeiss as well, if there's anyone there who can remember
what an LMS 510 looks like. You can certainly use Aviovision to tile
through a complete section or slide using a raster scan. Setting it up and
running is very easy [largely click beginning and end].

Being a precision XY stage and all, the Axiovision image sequence doesn't
need to 'stitch' the images together as such, as each tile will line up
perfectly with the next one [in all probability]. You can see it's a tiled
montage though [this conservative approach avoids image modification by
'stitching', and the tiling effect is of no scientific importance]. The
MetaMorph montage app works in a similar way. We [at UCL] used Axiovision
with a standard Axiovert 200M fluorescence/brightfield microscope, Zeiss
MRC full colour and Zeiss MRC B&W cooled fluorescence cameras, and Zeiss
[PALM] motorised stage.

Can't answer the main question though, as our LMS 510 here has the manual
XY stage - although I'll discuss this with out rep as we are thinking of
upgrading to a motorised stage some-time.

Those here at the WTCHG who have wanted tiled scans of entire sections
have generally preferred using a standard wide-field microscope over a
confocal - as transmitted light images acquired via camera generally have
higher resolution/image quality, and colour dye stained sections naturally
need a colour camera.

It should be easy to add cameras and Axiovision to the LMS 510 confocal
microscope [don't know your microscope though] and LSM PC, although you
won't get much, if any, change from £10,000 incl. camera hardware. We did
a similar thing, although the Axiovision & Zeiss MRC cameras were added to
our PALM micro-dissection laser system, rather than our LMS 510 [the
snazzy X,Y and diagonal PALM stage probably causes more problems than your
standard Zeiss one when talking to Axiovision].

I left the PALM behind 2 years ago, so things have probably moved on
Axiovision wise [although Zeiss is also quite conservative about about
updating software, tending to value software stability over a raft of new
features].

Regards

Keith

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

Dr Keith J Morris
Molecular Cytogenetics and Microscopy Core
The Wellcome Trust Centre for Human Genetics
Roosevelt Drive
Oxford
OX3 7BN
United Kingdom

Tel:  +44   ( 0 ) 1865  287568
Email:   [hidden email]
HomePage:  http://www.well.ox.ac.uk/cytogenetics

> Hi Gareth,
>
> As long as you have a motorized stage from Zeiss you should be able to
collect tiled images as any array combination as you want, and the stage
move]ing range is the limit. But you have to use a third party software
to
> stitch them together. I wrote a Matlab code to do so. I had done that
two
> years ago at St. Jude Children's Research Hospital, and now forget the
exact

> steps, but it was quite simple. The Zeiss software we used was LSM4.3.
>
> Lingqing
> **********************************
> Lingqing Zhang, PhD
> Nonlinear Optical Imaging Lab
> LM/EM Imaging Core Facility
> West Virginia University
> 1 Medical Center Drive
> Morgantown, WV 26506
> Phone: [phone]
> Email: [hidden email]
>
>
> On Fri, Feb 19, 2010 at 6:32 AM, Gareth Howell
> <[hidden email]>wrote:
>
>> Dear all
>> I have two questions relating to collecting tiles images on a Zeiss

510:
>> 1) I have a user who wants to scan large areas of tissue using our
confocal.
>> They want images at 1024 for good resolution. At the moment we are
using
>> the
>> Zeiss Stage Control tiling option but are limited to 4Kx4K pixel
arrays.
>> I
>> was wondering if someone had a protocol for using the MultiTime Macro
within
>> LSM510 software to collect multiple panels of 4Kx4K pixels with the aim of
>> stiching them all together for one big image.
>> 2) I'm probably missing a trick here but why can't I collect tiled images
>> using the camera within the LSM510 software; would it be possible with
AxioVision on the same microcope?
>> Regards
>> Gareth
>