luchen29 / CNN-Assisted-Parkinson-s-Disease-Classification

An end-to-end 3D convolutional neural network for Parkinson’s Disease diagnostic by using Tensorflow framework.

CNN-Assisted-Parkinson-s-Disease-Classification

Instructions:

The final code for 3d implementation are ‘learning_network_local.py’ and ‘input_3Dimage_local.py’. The final code for 2d implementation are ‘2d_cnn_bn.py’ and ‘dataset.py’.

Before running the code, please download the dataset. The Data are compressed as ‘GM’ and ‘WM’ two folders. Please download the two folders, use ‘train_test_split.py’ and ‘pd_hc_seperate.py’ to split the dataset.

Besides, please download the med2img from the following link: https://github.com/FNNDSC/med2image And run the code:)