Brightness level and heterogeity within set areas of a large number of images

I would like to ask for some help and advice on making measurements from images. The white in images is gases flowing into and out of a model of the airway. There are 3 areas where I would like to measure degree of ‘brightness’ and ‘heterogeneity’ (how much light and dark) within each area.

I have made standard image selection areas (see below)

Here are three examples of how the images may look like depending on how much gas within the model airway:

I have used the median and standard deviation of brightness to quantify the amount of “white” in the different selected area:


I have a total of 1200 images which makes for 3600 set of measurements. I would like a way of having:
count, mean, stdDev, min, max and mode, as well as the raw histogram (which I get from pressing copy) in a spreadsheet, ideally, one sheet per frame with three sets of measurements per sheet and most ideally, image filename also put within the spreadsheet. Would be great to be able to batch process as doing this 3600 times manually is not an exciting thought.

Is there a way of doing this and anyone who may be willing to help? Also, do you think the way I am looking at brightness and mixing (mixture of grey and white means mixing of gas) the best variables to use?

Many thanks

Reza

I’m not sure about getting the histogram Image but you should be able to write a macro to do this… This I’ve found can be a little overwhelming to get into at first, I’m still very new at this. But from what I’ve seen is really worth the time spent learning. There are several tutorials around the web that walk you through it… I’m not sure I’ve found one that is well suited entirely for someone new at coding etc.

Some of the basics in making loops etc. seem like a step learning curve but once you have them they can be applied to many Image analysis problems. This lists the commands etc. http://imagej.net/developer/macro/functions.html and opening the plugin macro recorder and doing one manually will provide much of the code

Check out the Segmentation page for an overview on how to do this sort of thing. Your goal is first to isolate the areas of interest—this is the segmentation step—and only afterward to perform the measurements.

That said, these images look quite challenging to segment. You could try selecting the areas of interest by hand in one sample, and posting the resulting mask here, so that we have something to work against.