Cortical porosity from slide

I’m trying to estimate cortical porosity from a H&E slide. I’ve subtracted background and made binary, then inverted to make the pores black and the bone white. Was it correct to invert the binary? I’ve tried running particle analyser in BoneJ, but it is measuring the osteocytes as well as the pores. I’ve set a scale using measurements from a micrometer. When I set a minimum volume, even if its ridiculously low, I get no results. If I do volume fraction it takes into account the osteocyte lacunae.

I’ve tried playing around with a few of the suggestions in this forum but haven’t yet found a solution. I’d love some thoughts on the best way I can get this to work. I couldn’t find the ROI interpolator .

If your image is 2D, you may be better off using ImageJ’s inbuilt Analyze Particles… (in the Analyze menu). There you can filter out particles based on area. Your Haversian canals look noisy - it may be helpful to filter you images with e.g. a median or Gaussian filter prior to binarising. Regarding the binary inverting, as long as your particles are in the foreground (pixel value 255), you are OK.

This is a more philosophical point: pores in bone exist at several length scales. Osteocyte lacunae (and the much smaller canaliculi) are pores, so are Haversian canals, so are the marrow spaces between trabeculae. Under EM, there are spaces within the matrix too where the organic and inorganic components interact. None of these pores are empty space. It’s important to define at the outset what you mean by ‘pore’.

Thanks Michael. Should have touched back to say I worked it out. I needed to split channels after subtracting the background, and set the scale using pixels not metric. Yes, ended up using particle analyser instead of BoneJ. Worked great until I accidentally closed down the whole program including my results summary instead of just the image I was working on when I had about 7 images (out of 40+) left to analyse. Thought I’d been saving it as I’d gone, but turns out not. Rookie mistake.

I hear your philosophical point, but we’d made a conscious decision to exclude lacunae :slight_smile:

1 Like