We've been looking at counting amyloid plaques in mouse hippocampus using 20x fluorescence images (antibody D5452). However when using either ImageJ or Matlab code, the fill holes or watershed options have thus far not worked to account for these dense core areas or partitioning plaques. I was wondering if anyone had further suggestions?
Hello Ashley,
You can try cellprofiler. As I understood, it is more about how you do segmentation to your images. Cellprofiler has many options to do so.
Hello Ashley,
as mentioned by Grigorii, it is a segmentation problem. You can try different automated thresholding algorithms (Otsu, Max Entropy, ...) and if non of them work well enough, you can try trainable segmentation in Image J (Weka or Labkit).
Thank you both! Labkit thus far seems to be working the best for a trainable segmentation.
- Badges
- Science method
- Similar topics
- Biological Science
- Histology
- Staining