Dear Jorge,

I have looked very closely at the paper by Kozlowski C, Weimer RM (2012) and it is certainly a fascinating and unique approach to quantifying microglial morphology. I think that it is worthwhile pointing out, however, that the parameters that this type of approach can yield are relatively impoverished to what can be obtained from digital reconstructions.

We have made extensive use of MBF biosciences Neurolucida package. In the base package you simply get the ability to trace images either online with a microscope or off-line with an image. It does have some nice features in that you can control branch thickness in a intuitive way and the information that you can is relatively easy to access and process. It is very time-consuming probably 10-30 min/cell. MBF also has an automated add on package that they call AutoNeuron which as its name suggests has been designed for neurons. We found that it does a reasonably good job of undertaking automated reconstructions of Microglia with that said it is not that "automated" (in realty it is semi-automated at best). To elaborate the issue, If you're familiar with microglial images obtained through standard fluorescent IHC you understand that you get many orphaned sectioned branches being labeled from sections above or below the plane of interest. When the Autoneuron package is doing its job it often mistakenly joins these branches. Creating a situation where you have to manually remove incorrectly corrected branches. The package does have settings that allow you to control what they refer to as 'gap tolerance' but this is a bit of a hit and miss affair. The downside with all MBF kit is that it is stupidly expensive (+10,000 AUD) and will only work on a limited number of microscope configurations. The only reason for using the automated package over the manual trace option is that it does a much better job at rendering the cell fully in three dimensions. This is something that is nearly impossible to do using "by eye" manual traces. We have found that two-dimensional traces to a reasonably well correlated with the three-dimensional traces so unless you're a fanatic the benefit might be questionable.

We have also looked at packages like Fiji which are relatively capable for undertaking analyses like skeletonization approaches but you'll need to be working with individual cells. The image processing routines are not sufficiently sophisticated to understand where one cell stops and and another cell starts. If you're interested I have just written a short review on this and would be happy to send it to you if you drop me your e-mail address and if you have any other specific questions at be happy for you to contact me.

Regards

RW

I and others have used fractal analysis to quantify microglial morphology using FracLac for ImageJ. Lacunarity and the fractal dimension of binary contours have proved especially useful in distinguishing among activation states, and in differentiating subtle differences, especially. The pattern analysis itself runs automatically and requires only brief training, but relies on extracting patterns from binary images up front. The ease of this depends on the starting images. If you can provide a few sample images, I might be able to put together an IJ plugin or macro to automate the pattern extraction process for you. I would also be glad to advise you on interpreting the analysis if you need any help at all.

Regards,

Audrey

Hi, thanks a lot for your answers. I will see what can I do with the material I have here.

Many thanks,

MBF StereoInvestigator

MBF StereoInvestigator and your vision are best.

Open source package: http://farsight-toolkit.org/wiki/L_Measure_functions

Computes basic features for swc file formats. Please let me know of any other desired feature that is not included and if you would like to use this.

I use MBF Neurolucida to trace the cell bodies and the processes. This will give you the area/perimeter measurements of the microglia as well as the process length and average process length. They key is to use sections thick enough so that you are not cutting off portions of the cell body or missing processes as they are in a different slice of tissue.

Dear Audrey,

thanks for your valuable comment! We have the same question and I would be glad to learn more about from you.

Many thanks, AS