Skip to main content
McNamara 20210126U INVERT_CONTRAST Olympus cellSens Dimension 3p1 - Process Manager - has its own file path - Kevin Murphy found 20210125M.png
(2021)
  • George McNamara
Image
Description
Olympus cellSens 3.1 tip as annotated screen shot for setting the folder paths to do large tile scanning --> image stitching. Need to optimize both "Tools - Options" (dialog near bottom left) and Process Manager (dialogs at center).
Need to set BOTH for EACH user (Windows 10 login).
To be successful at tile scanning, need fast PC, lots of ram (our current PC has 256 GB), fast drive (we now use NVMe M.2 SSD array for data and these temp folders), fast Ethernet card (now 10Gbe, though currently upload to user's JHU Microsoft OneDrive goes through 1 Gbe I.T. closet).
Our cellSens runs our "FISHscope", for details, see
** Note: our thanks to NIH's NIDDK P30 grant supplement and Prof. Mark Donowitz's internal JHU G.I. funding to purchase this microscope - specifications by Image Core Director Prof. Bin Wu and I, and help from John Gibas, Olympus sales; my thanks to Kevin Murphy, PhD, JHU Cardiology postdoc, for huge help optimizing FISHscope.
One of the major uses of FISHscope is single molecule RNA FISH (smFISH). We've imaged probe sets generated by "RNAscope" (expensive, can do 12plex in 3 cycles) and Bin Wu's lab (including 3 fluorophores --> 6plex by single and double labeling [A,B,C,AB,AC,BC], could "do" 10plex with four fluorophores, again, only single and double labels ... all doable because most RNA molecules are "sparse" in any cell ... Ialso note that ASI's spectral karyotyping (SKY) has used 5 fluorophores --> all 24 human chromosomes, Schrock 1996 Science ... metaphase chromosome spreads, so 2D sparseness).
A nice example of smFISH - using commercial probe set (48 oligos, 20 bases each, 2 base or larger gap) is
I was using smFISH probe sets from Biosearch Technologies in Houston, before moving to JHU.
Tile scanning and stitching enables efficient acquisition, since one user has mastered the approach, they can click Start and walk away (cellSens acquisition time notifier in status bar at bottom left is optimistic). Of course if you acquire a 100 GB dataset, you should plan to have a fast PC(s) in your lab to deal with your "data deluge" (we can discuss direct network/server access to transfer files, since OneDrive is slower).
Stage Navigator (bottom left) and "XY" (MIA" (bottom right) work together. Their Position List facilitates cyclic imaging aka Relocation because you can save the position list and load it at a later session. This is (or will be) powerful for "cyclic imaging", ex: 12plex (4x3) RNAscope, multiplex immunofluorescence, multiplex cyclic DNA-PAINT (single molecule and/or "macro" immunofluorescence).
Example: our current PC was able to acquire 48 tiles * 100 planes * 5 channels ("5plex") of 2048x2048 pixel, 16-bit images = 215 GB image file (at ~1 GB/sec file saving on our NVMe SSD array, 215 seconds = 3+ minutes in theory, in reality more like 15 minutes, probably due to cellSens * Windows ... a new generation PCIE gen 4 PC with lots of RAM would help a lot).


Keywords
  • tile scan,
  • stitching,
  • cellSens,
  • relocation,
  • cyclic labeling-imaging
Publication Date
Winter January 27, 2021
Comments
See description.
Citation Information
cellSens 3.1 settings for large datasets
Creative Commons license
Creative Commons License
This work is licensed under a Creative Commons CC_BY International License.