vem-primer-examples

Data analysis examples for the vEM primer (under development).

Examples

Here, we show two example data analysis workflows for volume electron microscopy.

• In the neurons example we segment neural processes in a small TEM volume of fruit-fly brain tissue, which is provided by the CREMI segmentation challenge. Here, we use the ilastik neural network and multicut workflow for segmentation.
• In the mitochondria example we segment and analyse mitochondria in a SBEM volume of human cortex, which is provided by the MitoEM segmentation challenge. Here, we segment and analyse the data using python notebooks.

In order to run these examples you will need to install a recent version (1.4 or newer) of ilastik for the neurons example. The python notebooks can either be run on google colab (links are available in the example folders) or on your own system using conda (see below). We also present some alternative analysis approaches for both examples.

Setting up conda environments (Advanced)

You can install either environment_cpu.yaml (if your machine does not have a GPU) or environment_gpu.yaml (if your machine has a GPU) via:

conda env create -f environment_cpu.yaml

or

conda env create -f environment_gpu.yaml

Then activate it via

conda activate vem-primer-cpu

or

conda activate vem-primer-gpu

You can then start a notebook server via jupyter notebook.

Note that you will need to update the CUDA version in environment_gpu.yaml according to your system specifications.