#Skeleton thickness and porosity_Fiji

##Orientation of the data set

Open data and reslice: image->stack->reslice

Reslice from the top or from right to make the stack horizontal

Save Substack by using ROI of intrast. ROI can be saved to ROImanager and reopened from there: Analyze->Tools->ROImanager

##Data segmentation using trainable weka segmentation

In Fiji, open Plugin -> Segmentation -> trainable weka segmentation

After open rename the labels of the class in setting.

Using the freehand tool, mark the area at each section and add to the relevant class. Repeat in about 20 slices.

Once the segmentation is complete, use the train classifier option. It is possible to save data, save classifiers, and reload them in the next set.

Make sure the segmentation is correct before creating the results.

After creating the result, adjust the threshold to BW using the Otso. Make sure the void value is 0 and the skeleton is 255, if not, invert the image: edit -> invert.

##Thickness calculation using Bone J

Reslice the binary stack from the bottom to get stack in the same orientation, and that slice one will be the top of the coral.

Open Fiji and make sure you have the BoneJ plugin; if not, add it.

###Fiji Open binary data set the scale at: analyze->set scale in micron

Open: Plugin->bonej->slicegeometry and set the parameters:

Plot the Mean Thick 3D (µm) by slice from the csv file

##Thickness and porosity using Python

On Coral1: open command prompt (CMD) Run: C:\ProgramData\miniconda3\Scripts\activate.bat Create a working conda environment: as instructed [here] (https://youtu.be/hDGSZMLS5F4?t=342)

After creating a working environment, install Python on it as instructed [here] (https://youtu.be/hDGSZMLS5F4?t=342)

Activate the environment by: conda activate XXX

Copy the path of the Python code and run it

The Python code works well on the Lynx system; however, on Windows, saving the calculated volume fails to save the numbers correctly. Therefore, it is possible to calculate the total volume with Fiji.

###To measure the total volume in Fiji, do this:

1. Open the total volume dataset
2. Image > Stacks > Z project, Projection type: Sum. It will produce a new image as output.
3. Analyze > Set measurements > check the ‘Integrated density’ box
4. Click on the result of the Z sum projection, then Analyze > Measure
5. The value in the last column, ‘RawIntDen’ is your volume in units of pixels
6. Multiply this by 151515 to get volume in um^3