DIPMP_exercises

Exercises of the course "Digital Image Processing in Medical Physics"

Description

The project contains the Python scripts and the input data for the exercises of the course "Digital Image Processing in Medical Physics". Each of the exercises is contained in a specific folder, with the name corresponding to the assigned exercise sheet. A test folder and a tutorial on digital image processing are also provided.

Connection

You can connect your machine to this GitLab repository and easily clone the code (described below for Linux). Alternatively you can simply download the .zip file

1. If you wish to use Linux, from a Windows or Mac, you can connect to the servers in the LMU physics CIP pool via X2Go: https://www.en.it.physik.uni-muenchen.de/dienste/netzwerk/rechnerzugriff/zugriff3/remote_login/index.html

2. You will need to connect to the VPN of the LRZ: https://www.en.it.physik.uni-muenchen.de/dienste/netzwerk/extern/index.html

Pleaes note that the Cisco VPN service (anyConnect) is no longer supported by LRZ, and will be shut down on June 2022. You can change to eduVPN to connect via VPN to the Munich university network. You can download the client here:

https://www.eduvpn.org/

The configuration is explained here:

https://doku.lrz.de/pages/viewpage.action?pageId=87425051&showLanguage=en_GB

Connect to the GitLab server

1. Login to your gitlab account with your LMU identification: https://gitlab.physik.uni-muenchen.de/

(see https://www.en.it.physik.uni-muenchen.de/dienste/datenspeicherung/gitlab/index.html)

2. Create your SSH key, see instructions at:

https://gitlab.physik.uni-muenchen.de/help/ssh/index.md

3. Navigate to User Preferences > SSH Keys

4. Copy your public key starting with ssh-rsa... and add it to your GitLab account

Clone the repository

1. On your machine, set up a directory for the exercises, e.g.:

   mkdir /path/to/exercises_folder/

2. Within this directory, clone the gitlab repository with (only possible with the SSH Key):

   git clone https://gitlab.physik.uni-muenchen.de/Marco.Riboldi/dipmp_exercises.git

3. You can also copy the repository by downloading a compressed file at the following public address:

https://gitlab.physik.uni-muenchen.de/Marco.Riboldi/dipmp_exercises

Computational framework

The code will be presented on the PyCharm IDE: https://www.jetbrains.com/pycharm/

1. On the CIP servers, PyCharm is installed. In the terminal, you can launch it via:

   pycharm_community_edition

2. Local installation on your PC:

https://www.jetbrains.com/help/pycharm/python.html

Use python = 3.9 (not all libraries supported yet for 3.10)

Create your project

Create your project using the following steps:

1. File -> New Project

Location: select folder (directory for the exercises)

2. Select New environment using: Virtualenv (set python 3.9 as base interpreter)

Location: select folder/venv

3. Then press Create

4. Warning that Directory is not empty: -> Create from existing sources

Virtual environment

The file requirements.txt lists all necessary python libraries and their versions used throughout the exercises

We recommend installing them within a virtual environment

https://www.jetbrains.com/help/pycharm/creating-virtual-environment.html#env-requirements

Once the project has been created, opening any .py file or requirements.txt prompts to the installation of the necessary packages

Click -> install requirements

The required packages can also be installed from the terminal:

pip install -r requirements.txt

If prompted upgrade your pip installation:

python.exe -m pip install --upgrade pip

Troubleshooting

Common issues for you local installation (Windows/MacOS):

1. Building wheel for scikit-learn (or simpleITK) � downgrade to python 3.8 as base interpreter

2. Issues with progressbar (Object not callable): � Uninstall and re-install progressbar2

Test run

Run test.py in the "test" folder. Type:

python \test\test.py

In the PyCharm terminal (assuming you are in the main repository folder.

Alternatively, you need to open the file test.py and define the run/debug configuration:

1. Define a name for your configuration

2. Select the virtual environment as the Interpreter

3. Complete the script path with the same path where you have the script test.py

4. You should see a CT image loaded and displayed in an output window