update README

README.md

@@ -1,13 +1,34 @@
 # PStrip
 This repository contains the code for the following paper (in writing)
 
+We've add a pre-release version of the code, as well as the sample atlases and pca_basis brain images. You can copy the data folder into the PStrip folder.
+
+We also have a dockerfile if you want to use docker. The dockerfile has all the setup commands for ubuntu16.04 with cuda 9.0
+
+If you are not using docker, you may also need to update the niftyreg install location.   
+
+Go to /PStrip/func_code, and change the nifty_bin in the niftyreg.conf file. The default location works for docker.  
+
+Once done, you may go to PStrip/main_code and run 
+```bash
+python pstrip.py --help
+```  
+to view all the options.    
+Or you may run   
+```bash
+python pstrip.py -i ../data/atlas/atlas_w_skull.nii  
+```
+which simply tries to extract the brain from atlas image.  
+
 ## Setup
 This code has been tested on Ubuntu 14.04 with Python 2.7 (Nvidia Titan X GPU, cuda 8.0)
 
 2017/10/19: CPU implementation is added. (No need for cuda, if only cpu is used)
 
 2017/10/06: the code works on Ubuntu 16.04 with Python 2.7 (cuda 9.0)
 
+This setup explains the procedure for PStrip. The exact command can be found in the Dockerfile.   
+
 **Download and Install virtualenv (Option for no-sudo privilege user)**
 Download virtualenv
 ```bash
@@ -41,10 +62,7 @@ CPU implementation of the decompostion is relatively slow, and we haven't upload
 **Download and Install niftyreg**
 http://cmictig.cs.ucl.ac.uk/wiki/index.php/NiftyReg_install
 
-For future purpose, n order to use the GPU, set the CUDA_FLAG to on. (Currently all registration steps are using CPU only)
-
-2017/10/17: We add a 'remove_nan' function to replace the nan value to 0, after affine.   
-2017/10/06: Newest niftyreg affine registration reg_aladin seems padding with NaN. It breaks the code at preprocessing.py.  
+For future purpose, in order to use the GPU, set the CUDA_FLAG to on. (Currently all registration steps are using CPU only)
 
 **Download and Install pycuda/scikit-cuda**
 Follow https://wiki.tiker.net/PyCuda/Installation/Linux