Summary and Schedule

This is a new lesson built with The Carpentries Workbench.

	Setup Instructions	Download files required for the lesson
Duration: 00h 00m	1. Course Introduction	How can I use this course to be better at my research?
Duration: 00h 10m	2. Medical Imaging Modalities	What are the common different types of diagnostic imaging? What sorts of computational challenges do they present?
Duration: 01h 10m	3. Working with MRI	What kinds of MRI are there? How are MRI data represented digitally? How should I organize and structure files for neuroimaging MRI data?
Duration: 02h 40m	4. Registration and Segmentation with SITK	What are SITK Images? How can registration be implemented in SITK? How can I segment an image in SITK?
Duration: 05h 10m	5. Preparing Images for Machine Learning	What are the fundamental steps in preparing images for machine learning? What techniques can be used to augment data? How should data from various sources or collected under different conditions be handled? How can we generate features for machine learning using radiomics or volumetrics?
Duration: 07h 15m	6. Working with Pathology Images	What are some open libraries can be used with pathology? How are pathology slides different from a photo? What is the structure of a pathology slide?
Duration: 08h 45m	7. Anonymizing Medical Images	What types of data make patient’s imaging data identifiable? How can I ensure the safe sharing of medical image data? How can I remove specific metadata from DICOM files?
Duration: 09h 55m	8. Generative AI in Medical Imaging	What is generative AI? How can generative AI be safely used in my work?
Duration: 10h 35m	Finish

The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.

Medical image analysis has become an essential tool in clinical research, enabling scientists to extract valuable insights from complex imaging data. This course is designed to enhance your computational skills in medical imaging, leveraging the power of Python to address advanced research questions. You already possess foundational skills in Python, and basic image processing, and this course will build on that knowledge to help you conduct sophisticated analyses with medical images.

The course begins with an overview of medical imaging and progresses to practical techniques for image processing using libraries like scikit-image, pydicom, and SimpleITK. You will then dive into feature extraction, machine learning applications, and statistical methods for analyzing anatomical differences. By the end of the course, you will have a comprehensive understanding of how to apply computational techniques to medical imaging research, enhancing your ability to conduct impactful studies and contribute to the advancement of clinical knowledge.

Installing the Python environment

---

We’ll use Mamba (a faster alternative to Conda fully compatible with it) to set up the Python environment. The environment YML file for the lesson is available on GitHub.

Installing Mamba via Miniforge

Miniforge is a minimal installer for Conda that includes Mamba. Here are the installation instructions for different operating systems:

Windows

1. Download the Miniforge3 installer for Windows from the official GitHub repository.
2. Run the installer and follow the prompts.

macOS and Linux

1. Open a terminal window.
2. Download the installer using curl or wget or your favorite program and run:

BASH

curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
bash Miniforge3-$(uname)-$(uname -m).sh

or

BASH

wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
bash Miniforge3-$(uname)-$(uname -m).sh

3. Follow the prompts to complete the installation.
4. Close and reopen your terminal to apply the changes.

If you encounter any issues, please refer to the official Mamba documentation.

If you’re unable to install Mamba, you can alternatively install Conda as a fallback option. If you choose this route, remember to replace mamba with conda in all subsequent commands in these instructions.

After you have a working Mamba (or Conda) installation you can proceed to create and activate the environment.

Creating the environment

1. Open your terminal:
   • On Windows: Open “Miniforge Prompt” or “Command Prompt”.
   • On macOS/Linux: Open your regular terminal.
2. Navigate to your lesson workspace directory.
3. Run one of the following commands (based on your preference for mamba or conda):

BASH

conda env create -f https://raw.githubusercontent.com/esciencecenter-digital-skills/medical-image-processing/main/learners/environment.yml

or

BASH

mamba env create -f https://raw.githubusercontent.com/esciencecenter-digital-skills/medical-image-processing/main/learners/environment.yml

4. Wait for the installation to complete. This may take a few minutes.

Activating the environment

After the installation is complete, activate the environment with conda or mamba:

BASH

conda activate medical_image_proc

or

BASH

mamba activate medical_image_proc

Additional environments

---

If you wish to run the code in the pathology section, you need to create a seperate environment for it. We have kept this environment seperate because our way to set it up involves using both conda and pip in the same environment. You can read the environment file and install by hand from all pip sources if you prefer an unmixed environment. Otherwise, follow the steps below:

1. Open your terminal:
   • On Windows: Open “Miniforge Prompt” or “Command Prompt” or “Git Bash”.
   • On macOS/Linux: Open your regular terminal.
2. Navigate to your lesson workspace directory.
3. Deactivate any current environment so you are in the base conda environment.
4. Run one of the following commands (based on your preference for mamba or conda):

BASH

conda env create -f https://raw.githubusercontent.com/esciencecenter-digital-skills/medical-image-processing/main/learners/environment_pathology.yml

or

BASH

mamba env create -f https://raw.githubusercontent.com/esciencecenter-digital-skills/medical-image-processing/main/learners/environment_pathology.yml

After the installation is complete, you can activate the environment with conda or mamba as follows:

BASH

conda activate pathy

or

BASH

mamba activate pathy

Downloading image files from Zenodo

---

Now that you have the environment set up, let’s download the necessary files from Zenodo.

Option A: manual download

1. Open your web browser and navigate to this lesson’s data Zenodo record address.
2. Look for the “Files” section on the page.
3. Click the download button for downloading the data.zip file containing the images.
4. Once downloaded, extract the contents of the ZIP file.
5. Move the extracted folder to your lesson workspace directory (where you’ll create notebooks and work during the lesson).

Option B: using Zenodo API

You can use the Zenodo API to download the files. The zenodo_get package should already be installed in your environment.

1. Ensure you’re in your lesson workspace directory and your medical_image_proc environment is activated.
2. Use the following command in your terminal (Miniforge Prompt if you are a Windows user):

BASH

zenodo_get 13311246

3. Extract the contents of data.zip.

Remember to keep the medical_image_proc environment activated throughout the entire lesson. If you close your terminal or restart your computer, you’ll need to activate the environment again using the activation command above.

Also, remember to verify that the extracted folder is in your lesson workspace directory (where you’ll create notebooks and work during the lesson).