Re: NeuronJ

[Erik Meijering]

Dear NeuronJ users,

Just a short note to let you know that I released a new update of NeuronJ
(version 1.3). The previous update (1.2) was released last July. Most
changes compared to last year's update (1.1) are really minor (I do have
lost of ideas for a major update but have no time to work on it currently...
sorry), but there are a few important improvements:

- Neurite centerline tracing in gray-scale (not 8-bit color) images showing
clipped intensities (overexpression) is now much better (previously, the
tracings were pushed away from the centerlines and towards the neurite edges
in these cases, because pixels with RGB value 0xFFFFFF corresponded to
background in the color images that I worked with, but this was a wrong
assumption in gray-scale images of course).

- I have added the possibility to export tracings as segmented line
selections (polyline ROIs), which enables you to import them as ROIs into
the images outside of NeuronJ and do all kinds of other measurements (such
as intensity profile plotting).

- The measurements dialog now also offers the possibility to do vertex
measurements (coordinates/values) and to show either calibrated or
noncalibrated results.

As usual, the latest version can be downloaded from:
http://www.imagescience.org/meijering/software/neuronj/

Please note that it requires ImageJ 1.38x and Java 1.4 or higher.

Best wishes,

Erik


On Wed, 27 Sep 2006 03:41:44 -0400, Erik Meijering wrote:

>Dear NeuronJ users,
>
>In response to several posts regarding pixel calibration:
>
>With the previous version of NeuronJ (1.0), the pixel calibration (pixel
>sizes and units) had to be set using the Parameters dialog of NeuronJ.
>
>With the latest version (1.1.0) of the program, these parameters can no
>longer be specified using the Parameters dialog. Since these are
>image-specific parameters, they can instead be set in the ImageJ > Image >
>Properties dialog. The values of these parameters stored in older NeuronJ
>data files are ignored.
>
>The latter version of NeuronJ also offers additional features. Please check
>out the NeuronJ webpage for more details:
>http://www.imagescience.org/meijering/software/neuronj/
>
>Best regards,
>
>Erik

[lechristophe]

That's great news. I just would like to know if you plan to make the
image auto-selection optional ? In NeuronJ 1.2, if you hover over the
image, it selects it automatically. For people like me who like to have
largest images view possible (for better compromise between zoom and
field of view) with various tollbars, logs, result tables etc...
overlaid on it, it's a real pain, because when going from a toolbar to
another it just put always the image in the front... Maybe this feature
is already just an option but I could'nt find how to disable it.

Thanks a lot for this tremendous plugin,

Christophe Leterrier

Erik Meijering a écrit :

> Dear NeuronJ users,
>
> Just a short note to let you know that I released a new update of NeuronJ
> (version 1.3). The previous update (1.2) was released last July. Most
> changes compared to last year's update (1.1) are really minor (I do have
> lost of ideas for a major update but have no time to work on it currently...
> sorry), but there are a few important improvements:
>
> - Neurite centerline tracing in gray-scale (not 8-bit color) images showing
> clipped intensities (overexpression) is now much better (previously, the
> tracings were pushed away from the centerlines and towards the neurite edges
> in these cases, because pixels with RGB value 0xFFFFFF corresponded to
> background in the color images that I worked with, but this was a wrong
> assumption in gray-scale images of course).
>
> - I have added the possibility to export tracings as segmented line
> selections (polyline ROIs), which enables you to import them as ROIs into
> the images outside of NeuronJ and do all kinds of other measurements (such
> as intensity profile plotting).
>
> - The measurements dialog now also offers the possibility to do vertex
> measurements (coordinates/values) and to show either calibrated or
> noncalibrated results.
>
> As usual, the latest version can be downloaded from:
> http://www.imagescience.org/meijering/software/neuronj/
>
> Please note that it requires ImageJ 1.38x and Java 1.4 or higher.
>
> Best wishes,
>
> Erik
>
>
> On Wed, 27 Sep 2006 03:41:44 -0400, Erik Meijering wrote:
>  
>> Dear NeuronJ users,
>>
>> In response to several posts regarding pixel calibration:
>>
>> With the previous version of NeuronJ (1.0), the pixel calibration (pixel
>> sizes and units) had to be set using the Parameters dialog of NeuronJ.
>>
>> With the latest version (1.1.0) of the program, these parameters can no
>> longer be specified using the Parameters dialog. Since these are
>> image-specific parameters, they can instead be set in the ImageJ > Image >
>> Properties dialog. The values of these parameters stored in older NeuronJ
>> data files are ignored.
>>
>> The latter version of NeuronJ also offers additional features. Please check
>> out the NeuronJ webpage for more details:
>> http://www.imagescience.org/meijering/software/neuronj/
>>
>> Best regards,
>>
>> Erik
>>    

[Bob Nienhuis]

Is there a way to trace in 3D in a stack? I want to trace some neurons in
thick Golgi stained sections.

Bob Nienhuis


On Nov 12, 2007 3:20 AM, Erik Meijering <[hidden email]> wrote:

> Dear NeuronJ users,
>
> Just a short note to let you know that I released a new update of NeuronJ
> (version 1.3). The previous update (1.2) was released last July. Most
> changes compared to last year's update (1.1) are really minor (I do have
> lost of ideas for a major update but have no time to work on it currently...
> sorry), but there are a few important improvements:
>
> - Neurite centerline tracing in gray-scale (not 8-bit color) images showing
> clipped intensities (overexpression) is now much better (previously, the
> tracings were pushed away from the centerlines and towards the neurite edges
> in these cases, because pixels with RGB value 0xFFFFFF corresponded to
> background in the color images that I worked with, but this was a wrong
> assumption in gray-scale images of course).
>
> - I have added the possibility to export tracings as segmented line
> selections (polyline ROIs), which enables you to import them as ROIs into
> the images outside of NeuronJ and do all kinds of other measurements (such
> as intensity profile plotting).
>
> - The measurements dialog now also offers the possibility to do vertex
> measurements (coordinates/values) and to show either calibrated or
> noncalibrated results.
>
> As usual, the latest version can be downloaded from:
> http://www.imagescience.org/meijering/software/neuronj/
>
> Please note that it requires ImageJ 1.38x and Java 1.4 or higher.
>
> Best wishes,
>
> Erik
>
>
> On Wed, 27 Sep 2006 03:41:44 -0400, Erik Meijering wrote:
> >Dear NeuronJ users,
> >
> >In response to several posts regarding pixel calibration:
> >
> >With the previous version of NeuronJ (1.0), the pixel calibration (pixel
> >sizes and units) had to be set using the Parameters dialog of NeuronJ.
> >
> >With the latest version (1.1.0) of the program, these parameters can no
> >longer be specified using the Parameters dialog. Since these are
> >image-specific parameters, they can instead be set in the ImageJ > Image >
> >Properties dialog. The values of these parameters stored in older NeuronJ
> >data files are ignored.
> >
> >The latter version of NeuronJ also offers additional features. Please check
> >out the NeuronJ webpage for more details:
> >http://www.imagescience.org/meijering/software/neuronj/
> >
> >Best regards,
> >
> >Erik
>

[Erik Meijering]

The option requested by Christophe is part of NeuronJ 1.3.1 released today.

Best regards,

Erik


On Mon, 12 Nov 2007 12:33:55 +0100, Christophe Leterrier
<[hidden email]> wrote:

>That's great news. I just would like to know if you plan to make the
>image auto-selection optional ? In NeuronJ 1.2, if you hover over the
>image, it selects it automatically. For people like me who like to have
>largest images view possible (for better compromise between zoom and
>field of view) with various tollbars, logs, result tables etc...
>overlaid on it, it's a real pain, because when going from a toolbar to
>another it just put always the image in the front... Maybe this feature
>is already just an option but I could'nt find how to disable it.
>
>Thanks a lot for this tremendous plugin,
>
>Christophe Leterrier

[Erik Meijering]

Dear Bob,

NeuronJ was developed for doing 2D work only. Bitplane independently
developed a 3D interactive neuron tracing tool (the FilamentTracer module of
Imaris) around the same time NeuronJ was first released, taking away the
incentive for me to develop such a tool (please understand that in my lab
our job is to develop new image analysis methodology, not software tools per
se - they are spin-off products of the scientific and experimental work we
do). Maybe some time in the future, in the context of new research projects
in neuroscience, we will develop a 3D version of NeuronJ after all... but at
this point I cannot promise anything.

Best regards,

Erik


On Tue, 13 Nov 2007 10:49:30 -0800, Bob Nienhuis wrote:
>Is there a way to trace in 3D in a stack? I want to trace some neurons in
>thick Golgi stained sections.
>
>Bob Nienhuis

[Mark Longair-2]

Bob Nienhuis <[hidden email]> wrote:
>
> Is there a way to trace in 3D in a stack? I want to trace some
> neurons in thick Golgi stained sections.
>
> Bob Nienhuis

Bob,

If you would like to try an alternative semi-automatic neuron
tracer for 3D stacks, you could try this plugin:

  http://homepages.inf.ed.ac.uk/s9808248/imagej/tracer/

Very few people have tested this so far, so I would be grateful
for any feedback (off the list).

regards,
mark