Mitochondria Segmentaion

This tutorial provides step-by-step guidance for mitochondria segmentation with the EM benchmark datasets released by Lucchi et al.. We consider the task as a semantic segmentation task and predict the mitochondria pixels with encoder-decoder ConvNets similar to the models used in affinity prediction in neuron segmentation. The evaluation of the mitochondria segmentation results is based on the F1 score and Intersection over Union (IoU).

All the scripts needed for this tutorial can be found at pytorch_connectomics/scripts/. Need to pass the argument --task 2 when executing the train.py and test.py scripts. The pytorch dataset class of synapses is torch_connectomics.data.dataset.MitoDataset.

1. Get the dataset:

   1. Download the dataset from our server:

      wget https://hp06.mindhackers.org/rhoana_product/dataset/lucchi.zip
      

   For description of the data please check the author page.

2. Run the training script. The training and inference script can take a list of volumes and conduct training/inference at the same time.

   $ module load cuda/9.0-fasrc02 cudnn/7.0_cuda9.0-fasrc01 boost # on Harvard rc cluster
   $ source activate py3_torch
   $ python -u train.py -i /path/to/Lucchi/ -din img/train_im.tif -dln label/train_label.tif -o outputs/unet_res_mito\
     -lr 1e-03 --iteration-total 60000 --iteration-save 10000 \
     -mi 112,112,112 -ma unet_residual_3d -mf 28,36,48,64,80 -me 0 -daz 1 -moc 1\
     -to 0 -lo 1 -wo 1 -g 4 -c 4 -b 4
   

   • Data: i/o/din/dln input folder/output folder/train volume/train label

   • Optimization: lr/iteration-total/iteration-save learning rate/total #iterations/#iterations to save

   • Model: mi/ma/mf/moc/me/daz input size/architecture/#filter/#output channel/with 2D embedding module/z-data-augmentation

   • Loss: to/lo/wo target option/loss option/weight option

   • System: g/c/b #GPU/#CPU/batch size

3. Visualize the training progress:

   $ tensorboard --logdir runs
   

4. Run inference on test image volumes (change LOG-FOLDER): VOC-test=0.945

   $ python -u test.py -i /path/to/Lucchi/ -din img/test_im.tif -o outputs/unetv0_mito/result\
     -mi 112,256,256  -g 1 -c 1 -b 1 -ma unet_residual -mf 28,36,48,64,80 -me 0 -moc 1
     -mpt outputs/unet_res_mito/LOG-FOLDER/volume_59999.pth -mpi 59999 -dp 8,64,64