= Tutorial=
{F8105850}
{F8105851}
{F14084374}
Outdated : F8105712
[Worflow and Data management](https://docs.google.com/presentation/d/1jfdDF_viFlqg3St1r2829cHj5HYXFfUfRe7wRWt3410)
= Relocate Settings =
{nav Settings > Relocate Settings}
{F23605161, size=full}
[[ https://c4science.ch/w/bioimaging_and_optics_platform_biop/image-processing/imagej_tools/perkinelmer-stitching/ | Using Fiji to process your Operetta images]]
= Mind the Byte =
**One full well from a 96-well plate **(4.5mm x 4.5mm field of view)
Full resolution acquisition **20x air objective**
**1Gb per plane **(Single channel, single slice, single timepoint)
Example acquisition, 2 channels, 11 slices : **9.35 GB per well**
= Write archive =
It takes **~1 hour for 100 GB**
= Harmony 4.8 installation =
**Start Autorun.exe**
`\\svfas6\biop\biop\Informatics\Software Library\Harmony\2019-10-17 SW-Update Harmony 4.9 MDe`
Select: Office PC > install
Warning : if the computer states : "you need to restart the computer", restarting the software will not do anything...
To proceed with the installation :
* Run Regedit.exe
* Find the Windows Registry : HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager : PendingFileRnameOperations
* Delete the PendingFileRnameOperations registry (right click > delete > ok)
* You can now proceed to do the installation
The **Region and Language Settings of the PC should be in English UK or US** otherwise the software crashes when you start it.
Connect the Operetta database.
Start Harmony 4.9.
Go on: settings > data management > database settings
Add a new database (click on **+**)
Computer name: SV-01-366
Alias name: Operetta CLS
= Plate reference =
| **Brand**| **Ref Name** | **Ref Nbr** | **Qty** | **EUR**| **Remarque** |
|PerkinElmer | [[ http://www.perkinelmer.com/fr/product/cellcarrier-96-ultra-lid-8x20b-6055300 | CellCarrier Ultra 96 ]] | 6055300|50 | 388.00 | works with 20x et 63x water|
|PerkinElmer | [[ http://www.perkinelmer.com/fr/product/cellcarrier-96-ultra-lid-8x20b-6055308 | CellCarrier Ultra 96 ]] | 6055308 |160 |1241.00 |
|PerkinElmer | [[ http://www.perkinelmer.com/fr/product/viewplate-96-f-tc-10x5b-6005182 | ViewPlate-96F]] | 6005182 | 50 | 390.00 | Can't access the wells on the border with high mag. obj. + issue with finding focus using 40x & 63x |
|PerkinElmer | [[ http://www.perkinelmer.com/fr/product/viewplate-96-f-tc-10x5b-6005225 | ViewPlate-96F]] | 6005225 | 5x10 | 372.00 |
|Falcon | [[ https://ecatalog.corning.com/life-sciences/b2c/EUOther/en/Microplates/Assay-Microplates/96-Well-Microplates/Falcon%C2%AE-96-well-Polystyrene-Microplates/p/353219 | 96Well ImagingPlate ]] | FA-353219 | 4 x 8 | CHF 206-290 | Can't access the wells on the border with high mag. obj. + issue with finding focus using 63x |
|Cellvis | [[https://ibl.at/ecommerce/product?category_id=150721&product_id=492523 | 12 Well Cellvis]] | 492523 | 1x20 | 265.00 Eur. | Works with 40x water and ? |
= Export flat-fielded images =
It is possible to export the flat-fielded image by using the script located in `\\svraw1\ptbiop\public\0-Software\Operetta\Export_Flatfielded_Images` inside the Harmony Software
This is a small tool which can apply the flat-field correction. It is based on the image processing core of Harmony (acapella.exe). It runs on Harmony 4.8 and seems to work fine. Should you still have Harmony 4.6 you probably have to edit the batch file for the correct path to acapella.exe. Unfortunately I can’t test it on older versions, but I don’t really expect problems.
- First you have to export measurements from Harmony. On the Operetta CLS this can even be exports with “Referenced Images”, i.e. you will just get an “index.ref.xml” file. Or if you prefer the normal export, the “index.idx.xml” is also fine.
- Then you run the batch file, which just calls the acapella.exe interpreter with the script file as parameter. This script will then ask you for an index.xml-file. Select the file that was created during the Harmony export.
- It will create a new folder and write all flatfield corrected images into that folder. It uses a similar naming convention as Harmony, just with the prefix “flex_”. It will also generate a new appropriate index.xml file, in case your system works with the xml files.
These scripts can run on any Harmony installation that has a valid license dongle.
# Tips and Tricks
## Digital Phase Contrast (DPC)
NOTE: When setting up the DPC parameters, find a field where there are no artifacts or floating dead cells as these will affect the optimization.
1. Find the focus position (offset) in brightfield where you see your cells the **least**. That is, the most in-focus plane.
2. Set your exposure time and laser power so that you are using a good part of the histogram (max values around 50000 gray levels
3. Set yourself on "High Contrast" Mode and hit "Snapshot DPC"
4. The system will acquire 50 images and find the ones with the highest contrast to use as top plane and lower plane
5. Set the DPC to "Manual"
6. Change the Filter and Speckle Scale in the following way": Speckle to 0 in order to keep mitotic cells, otherwise they get removed. Filter to 0.5 if your cells have holes in them, or larger than 1 if there is too much background around the cells.
## Viewing fused wells
When acquiring images in Test mode, you can select all of the fields, Right Click and select "Wells and Fields, Realistic".
This opens a fused version of the images in a separate window.
## Adding an overview image to the well grid.
After you acquire a test images of multiple tiles, Select the images and select "Use Background for Well"
The fused image will appear as the background now, to help you select fields of interest when setting up the Acquisition
# Using 35mm Petri Dishes
If you want to image a petri dish with glass bottom (No 1.5 - 170 um thickness), you can use the adapted X-wing holder designed by Oli. The layout to select is `petri dish 35mm`.
# Using ibidi slides
Please have a look at https://github.com/BIOP/tool-ibidi-slide-holder:
* Well Slides tested and used:
** https://ibidi.com/chamber-slides/44-3-well-chamber-removable.html
** https://ibidi.com/chamber-slides/45-8-well-chamber-removable.html
# Autofocus Failure: Debugging
In case there is a problem with the autofocus, you can call the Cockpit Data Plotter to see how it is trying to make the autofocus on the system.
{nav WIN > PerkinElmer > Service > type=instructions, name=FieldApp}
Password: `Scan4FAS`
{F21587771, size=full}
This plot shows the result for a "Two Peak" autofocus (In this case for the 40x Water objective, so from left to right: Water-Glass interface followed by Glass-Water. The sample would be on the right of this plot.
{F21587855, size=full}
Example of failed autofocus: The large peak on the right here represents the interface of a transwell that is inside the well. Because the system focuses on the "largest peaks" it is unable to find a similar second peak and thus fails.
## Option 1:switch to the "One Peak" mode.
In the example above, it would be useful as then the system would use the transwell as a reference (Because it is the strongest peak)
## Option 2: Change the geometry of the plate
If we change the geometry of the plate so that the system expects it to be **lower** than it actually is:
{F21587914}
In this case, if we change the value of **H** from 1.00 to 0.50, the Operetta will just stop scanning for reflections before finding the transwell, resulting in a successful Two Peak detection.
In the case of the CellVis Plates, two configurations were made for the Transwells.
{F21588074, size=full}
Use the **Two Peak AF** plate for standard plates without transwells
# Flatfield and Orientation Matrix
These two calibrations are computed **on the fly for each experiment** In case there is no Flatfield correction available, this is because there were not enough images that Harmony considered of sufficient quality to estimate a flatfield. This can happen when images touch the borders of the wells. The software will inform you after the experiment is run
{F21587813, size=full}
Top: Successful flatfield estimation from enough images (30 images in total)
Bottom: Unsuccessful flatfield,experiment only contained 4 images in total.