Syllabus

Duration: 3 Days + Optional Day 4

Learning Outcomes:

By the end of this course, participants will:

  1. Gain foundational skills in clinical scientific computing, including data pipelines, database queries, and GUI development.
  2. Develop practical knowledge of image processing and DICOM workflows.
  3. Optionally, acquire foundational knowledge in data science, machine learning, and AI for medical physics on Day 4.

Assessment

Participants will submit:

  1. Practical outputs: database queries, scripts, GUI, and DICOM tasks.
  2. Data science and AI exercises (optional for those attending Day 4).
  3. A case study on metadata transfer and reflections on clinical workflows.

Note:

Day 4 is optional and designed for participants interested in extending their knowledge of data science and AI applications in medical physics.

Day 1

Day 1: Core Processes and Database Fundamentals

Learning Outcomes Covered: S-BG-R1 #3, #7, #8, #10, DOPS; S-CSC-R1 #5, #6; S-CE-S2 #1

Morning Session: Observing Processes in Clinical Computing

  • Lecture Topics:
    • Overview of internal data-processing pipelines.
    • The software development lifecycle (SDLC) phases and their relevance to clinical practice.
    • Database systems and their role in clinical scientific computing.
  • Practical Tasks:
    • Observe a running internal data-processing pipeline and reflect on the process (S-BG-R1 #3).
    • Reflect on the SDLC in clinical contexts (S-BG-R1 #7).
    • Observe database information retrieval and external data pipelines (S-BG-R1 #8, #10).

Afternoon Session: Database Development and Queries

  • Lecture Topics:
    • Introduction to relational databases: designing tables, creating input forms, and writing queries.
  • Practical Tasks:
    • Write a database query to retrieve specific data (S-BG-R1 DOPS).
    • Develop a relational database with forms and queries (S-CE-S2 #1).
    • Download a dataset and transfer it to a clinical system (S-CSC-R1 DOPS).

Day 2

Day 2: Scripting and Advanced Development

Learning Outcomes Covered: S-CSC-R1 #7, #9, DOPS; S-CE-S2 #10, S-CE-S2; S-CE-S3 #9, S-CE-S3

Morning Session: Scripting for Data Processing

  • Lecture Topics:
    • Fundamentals of scripting for data processing: variables, loops, conditionals, and data structures.
    • Debugging and modifying software in clinical environments.
  • Practical Tasks:
    • Write a script demonstrating variables, loops, and pattern matching (S-CE-S2 #10).
    • Debug or modify an existing piece of software (S-CE-S2).

Afternoon Session: GUI/Web Development and SDLC in Practice

  • Lecture Topics:
    • Designing graphical user interfaces (GUIs) or webpages for clinical data presentation.
    • Reviewing quality management systems and legislative influences in clinical computing.
  • Practical Tasks:
    • Develop a simple GUI or webpage for presenting clinical data (S-CE-S3 #9).
    • Train users for a newly developed GUI solution (S-CE-S3).
    • Review the impact of legislative and quality standards on clinical computing practices (S-CSC-R1 #6, #9).

Day 3

Day 3: Clinical Image Processing and Security

Learning Outcomes Covered: S-IN-S1 #14–#19; S-NM-S4 #5–#10; S-DR-S3 #9; S-C1 DOPS

Morning Session: Image Processing and DICOM Workflows

  • Lecture Topics:
    • Basics of DICOM headers and metadata manipulation.
    • Techniques for image registration and segmentation, including AI-based methodologies.
  • Practical Tasks:
    • Write, validate, and document software to manipulate DICOM headers (S-IN-S1 #14).
    • Investigate methods for registering image datasets (S-IN-S1 #15).
    • Explore segmentation techniques using AI and traditional methods (S-IN-S1 #16).

Afternoon Session: Image Analysis, Security, and Integration

  • Lecture Topics:
    • Fundamentals of clinical image analysis: extracting quantitative data.
    • Ensuring confidentiality and security in clinical workflows.
  • Practical Tasks:
    • Perform basic image analysis using clinical tools (S-NM-S4 #6).
    • Extract quantitative data from image studies (S-NM-S4 #7).
    • Prepare a case study on metadata transfer in clinical imaging (S-NM-S4 #8).
    • Apply appropriate security methods to clinical datasets (S-C1 DOPS).

Day 4

Optional Day 4: Introduction to Data Science and AI for Medical Physics

Learning Outcomes Covered (Optional): Data Science and AI Skills

Morning Session: Fundamentals of Data Science and Machine Learning

  • Lecture Topics:

    • Introduction to data science and its role in medical physics.
    • Data types, structures, and tools in medical physics.
    • Basics of machine learning: supervised and unsupervised learning.
    • Applications of machine learning in medical physics: dose prediction, automated contouring.
  • Practical Tasks:

    • Load a mock dataset using pandas in Google Colab.
    • Perform basic data cleaning and create simple visualizations using Matplotlib.
    • Build a linear regression model for tumor growth prediction using scikit-learn.

Afternoon Session: Deep Learning and Advanced AI Applications

  • Lecture Topics:

    • Introduction to neural networks and deep learning.
    • Convolutional Neural Networks (CNNs) for medical imaging tasks.
    • Challenges and ethical considerations in AI for medical physics.
  • Practical Tasks:

    • Use TensorFlow to fine-tune a pre-trained CNN for tumor classification in Google Colab.
    • Apply transfer learning to medical datasets in Kaggle.
    • Evaluate the model and interpret results.