PAC18
sMRI based depression classification
Directory Structure
"./models" : for all saved models (in gitignore)
"./raw_data/pac2018_data" : for training data (in gitignore)
"./processed_data/hdf5_files" : for training data after pre processing if required (in gitignore)\
Packages Required
jupyter
pytorch
pandas
simple ITK
nibabel
About ./src/*
-
DataAugment : Contains all required functions for 3D data augmentations
-
DataGenerator : Contains function to load data batch wise..
-
DensenetModel : Contains 3D Densenet model.....
-
Trainer_Tester : Contains train and test class with all train and test strategy
Working....
-
First use Create_Data.ipynb to create prpcessed data
-
Use ./src/main.py to train/test model
Contact
- Avinash Kori (avinashgkori@smail.iitm.ac.in)
If any comments or issues, pull requests/issues are Welcomed....
Thankyou