Identification of "AIA terminal nodule" and "AIA central nodule" using DenseNetFCN-3D.

Based on two different 3D movies (20190805_SJR3.2.2_w1_s2.nd2 and 20190805_322_w1_s2.nd2), provided by the Laboratory of the Physics of Biological Systems, the task is to identify and distinguish two parts (the terminal nodule and the central nodule) of a neuron type called AIA.
3D frames of videos were used as isolated images for training, using a DenseNetFCN-3D with 4 dense blocks of 3 layers per block.

Deliverables:

• DataGenerator.py python file: Create generator for training and validation, user may apply image augmentation on the fly.

• DenseNet3D.py python file: Implementation of the DenseNetFCN-3D from GalDude33.

• generateFramesMasksFromVideo.ipynb notebook: Generate frames/masks with desired shape from the video/ground_truth files.

• Training.ipynb notebook: Feed the DenseNetFCN-3D with frames/masks pairs.

• VisualizationPrediction.ipynb notebook: Predict an example mask with the DenseNetFCN-3D model.

• ConfusionMatrix.ipynb notebook: Compute the accuracy and the F1 score for the 'AIA terminal nodule' (class 1) , the 'AIA central nodule'(class 2) and the other cells (class 0).

(keep all the files in the same directory)

How to train the DenseNetFCN-3D, predict masks and note the model:

1. The training samples and corresponding masks to be stored as .npy files, formattted as frame_i.npyand mask_i.npy respectively (i=frame_time). Specify PATHs in generateFramesMasksFromVideo.ipynb.
   P.S: Labeled ALL cell parts that were not AIA central/terminal noduls as same feature

2. Specify path and variable values in Training.ipynb notebook. Notebook trains a DenseNetFCN-3D and saves the model.

3. Visualization of performance with the VisualizationPrediction.ipynb notebook. The last cell shows the image, the mask and finally the predicted mask.

4. The ConfusionMatrix.ipynb notebook may be used to compute the accuracy and the F1 score for the 'AIA terminal nodule' (class 1) , the 'AIA central nodule'(class 2) and the other cells (class 0).

REFERENCES: DenseNet3D from GalDude33

Collaborators: Victor Stimpfling, Loris Pilotto