3D U-Net ovary and follicle detection, Ext1 and Ext2

Dependencies

The code requires the following libraries:

• tensorflow 2.x
• albumentations
• matplotlib
• numpy
• scipy
• scikit-image
• h5py
• nilearn
• nibabel

USOVA Dataset

The dataset is available here.

Network training (extension 1 or 2)

Training command:

python train_USOVA3D_unet_ext.py \
	-m model_ext1.h5 \
	-r 1e-3 \
	--numFilt 8 \
	--extensionType ext1 \
	-e 200 \
	--trainData trainSet.h5

will train a model with learing rate 1-e3, 8 filters on the first U-net level, with extension ext1. It will train for 200 epochs with the data loaded from trainSet.h5. The model will be saved to model_ext1.h5.

To evaluate on the training data:

python predictALL_USOVA3D_unet_ext.py \
	-m model_ext1.h5 \
	--outOvaries predictionsOvaries.h5 \
	--outFollicles predictionsFollicles.h5 \
	--testData testSet.h5

Publication

If you use this code or models in your publication please cite:

• B. Potočnik, M. Šavc, Deeply-Supervised 3D Convolutional Neural Networks for Automated Ovary and Follicle Detection from Ultrasound Volumes , Applied Sciences, 2022, submitted.