OME schema parameters

[SteveFirth]

Hi,
I am interested in using OME Tiff within my microscopy facility and potentially establishing OMERO. We use a range of different microscopes and OME is very attractive in consolidating formats.
The OME format seems to be missing several of the parameters that we need, but as a biologist reading these schema I may be not reading things correctly. Could people please inform me whether the following are indeed contained within the current schema, ticketed for future release, or missing and I should suggest their inclusion.

Variable band filter positions - both Olympus and Leica confocal microscopes use variable wavelength detection, not specified by a filter cube.
Wavelength analysis using multiple PMT's (Zeiss and Nikon spectral imaging).
AOTF/laser intensity
Relative Power on Ar gas laser/diode lasers
For stacks or multidimensional images - number of channels, z stacks, time points, total number of images
Field rotation
Accumulation and Averaging - line or Frame
Scan direction - uni or bi directional scanning

Cheers, and thanks for your time,
Stephen

[ajpatterson]

Hello Stephen.

Glad to hear you think the OME-TIFF will be useful to you.

> The OME format seems to be missing several of the parameters
> that we need, but as a biologist reading these schema I may
> be not reading things correctly.

And I am a computer scientist so have problems with the biology, the physics, oh lots of things

Here are my suggestions of where these might be stored - if it does not sound right get back to be as we keep finding new techniques we need to represent.

> Variable band filter positions - both Olympus and Leica
> confocal microscopes use variable wavelength detection,
> not specified by a filter cube.

I assume each channel is only using one position of the filter band.
In this case I would store it in the TransmittanceRange of the Filter, this Filter is then part of a FilterSet and the Channel for this band has a FilterSetRef pointing to the right FilterSet.
http://cvs.openmicroscopy.org.uk/svn/sp ... html#id189

> Wavelength analysis using multiple PMT's (Zeiss and Nikon spectral imaging).

Is each Photomultiplier Tube producing one set of pixel data for a single wavelength?
If this is the case then each is creating one Channel. Channel is a child of Pixels in the model.
http://cvs.openmicroscopy.org.uk/svn/sp ... html#id250

> AOTF/laser intensity

Acousto-Optical Tuneable Filter
We have a flag on Laser to say if it is Tuneable but nothing on Filter.
Is this Intensity a property of a Laser, or a property of a Filter, or both?
What units is it defined in?
Do you know of a commercial file format that stores it?

> Relative Power on Ar gas laser/diode lasers

Possible locations are:
- Lightsource has Power (milliWatts)
- LightSourceRef has Attenuation (%)

> For stacks or multidimensional images - number of channels,
> z stacks, time points, total number of images

These values are not explicitly stored. They can be calculated from the other data.

> Field rotation

Can you define what you mean?

> Accumulation and Averaging - line or Frame

Can you define what you mean?

> Scan direction - uni or bi directional scanning

This is not recorded at present, is this a property of the lightsource, or detector, or elsewhere?

Look forward to your responses,

Andrew

[SteveFirth]

Hi Andrew,
Thanks for the answers, here's some more questions and some definitions.

Wavelength specification on Leica and Olympus confocal- Transmittance Range, perfect, didn't see this parameter.
Spectral imaging on mutiple PMT (and wavelength scanning with Olympus and Leica) - all can be recorded as individual OME-Tiffs with wavelengths specified in Transmittance range.

AOTF - controls laser intensity, essentially a linear ND filter. Units are % transmission, the inverse of Attenuation in LightSourceRef.
Ar/diode laser variable power - these units have a power rating, e.g. 40mW, but then have a power control (%) dial.
So a 30mW Ar gas laser can be at 20% current and an AOTF of 10% transmission. Is it possible to put 30mW and 20% current into LightSourceRef-Power?

Field rotation - the field of view/capture can be rotated to frame an image better or for consistency (e.g. always imaging a zebrafish with head to the left and tail to the right of the image with spine horizontal). Not particularly important to record for single images but useful for stitched images and tracking.

Accumulation - a line or frame is scanned multiple times and the results are added together to intensify a dim image.
Averaging - the line or frame is accumulated then divided by the number of scans to reduce noise.

Uni or bi-directional scanning - confocal scanner generally operate on a single direction scan, only recording data on one direction of the raster scan. For faster scanning they can be made to record data on both directions of the raster scan, but this may affect image quality. When reviewing an image it would be good to know in what mode it was captured.

If any of these parameters is new and needs inclusion in the OME schema how does a 'ticket' get generated?

Thanks for your help,
Steve

[wmoore]

Spectral imaging on mutiple PMT (and wavelength scanning with Olympus and Leica) - all can be recorded as individual OME-Tiffs with wavelengths specified in Transmittance range.

This really requires a change in our model to support this properly. We have been discussing the support of dimensions in addition to XYZCT, E.g. wavelength for spectral scans.
Currently, the image could be represented in a single OME-TIFF with one plane per channel, and a different Emission filter (with it's own transmittance range) for each Channel.

AOTF - controls laser intensity, essentially a linear ND filter. Units are % transmission, the inverse of Attenuation in LightSourceRef.

Probably best to store this as Attenuation then, if that works for you?

Is it possible to put 30mW and 20% current into LightSourceRef-Power?

I'll note this in the ticket we have for LightSource http://www.ome-xml.org/ticket/111

Field rotation -

OK. How is this measured? Degrees? 0 - 180 or 360? - measuring from what to what? Is there a convention? Do any existing file formats include this parameter?

Accumulation - a line or frame is scanned multiple times and the results are added together to intensify a dim image.
Averaging - the line or frame is accumulated then divided by the number of scans to reduce noise.

These are attributes we are currently considering for Detector. Is Averaging simply a value like "Line" or "Frame" or "None" (False) ? Do you ever have Accumulation = 4 and Averaging = False, so that the pixel values are summed?

Uni or bi-directional scanning

OK.

how does a 'ticket' get generated?

Go to http://www.ome-xml.org/ and log in as "ome" "ome". Now you can create a new ticket.
Alternatively, let us know what you want and we can add or amend existing tickets etc.

Many thanks,

Will.

[wmoore]

Forgot to add: The detector changes ticket is here: http://www.ome-xml.org/ticket/65

Will.

[SteveFirth]

Hi Will,
thanks for your comments.
The spectral imaging change you suggest sounds good.
Attenuation for the AOTF is a good fit, attenuation =100-AOTF%
I'll have a read of the LightSource Ticket to see what changes are already on the way and the process for ticket generation and completion.
Field rotation is measured in degrees. The Leica LAS AF software (.lif files) include this, they refer to it as the scan field rotator and scan rotation is measured in degrees (-100 to +100).
Averaging and Accumulation can be performed independantly to either average the values of a pixel over multiple scans, or sum the pixels from multiple scans.
Averaging and Accumulation can both be performed by line or frame, there is a number of iterations associated with the parameter, e.g. Line average x 4. Both can be used at the same time, but one must be frame and the other line, it does not matter whether it is line accumulation or frame, just cannot have line averaging and line accumulation, but you can have line averaging and frame averaging at the same time and line accumulation and frame accumulation at the same time.
Cheers,
Steve

[wmoore]

Hi

I made ticket for Field Rotation: http://www.ome-xml.org/ticket/116
All the example LIF files I have contain a "Scan Rotation" Field, E.g. 77 degrees (or 0 degrees in most cases) and a "Rotation Direction" which is always "1" in all my examples. Is this value -1 if the rotation is the other direction? Is the "Scan Rotation" value always positive?

The Averaging and Accumulation is sounding complex! I see that LIF files have these attributes:

{{{
Frame-Accumulation 1
Frame-Average 4
Line-Average 1
}}}

Presumably Frame-Average is an integer and all the 1 / 0 are True / False (or is Line-Average an integer?) Are these attributes sufficient to describe what you want? Or is there a Line-Accumulation attribute too? Do you have an example file you can send (export as XML)?
If you have Line-Accumulation AND Frame-Accumulation, you are scanning each line several times per Frame scan, and then you are also performing multiple Frame scans?

We have to be careful that we don't make our model limited to one particular file format, and that it is applicable to the functionalities of all confocal scopes (or as many as possible).

Cheers,

Will.

[wmoore]

Hi,

I just found some more info on Line/Frame Average/Accumulation
http://www.hi.helsinki.fi/amu/AMU%20Cf_ ... verage.htm
so I think I understand a bit better now.

Looks like we need these attributes:
Frame-Accumulation
Frame-Average
Line-Accumulation
Line-Average
All of which are integers. This would allow every possible configuration, but might be a bit of an overkill!

Will.

[SteveFirth]

This would allow every possible configuration, but might be a bit of an overkill!

Will.

Hi Will,
as a manager of a microscopy facility I know someone will use any and all parameters that you provide.

How much extra space(memory?) does an extra parameter like this add to the XML?
Cheers,
Steve

[cxallan]

Very little overhead Steve. Really the only burden is modeling complexity and maintenance on our side.