Normative modelling educational course OHBM 2024 🧠🧠

Date, location and time :

Sunday, 23rd of June, 9am-1pm GMT+9, Seoul, South Korea.

Normative modelling has revolutionized our approach to identifying effects in neuroimaging, enabling a shift from group-level statistics to subject-level inferences. The use of extensive population data, now readily accessible, allows for a more comprehensive understanding of healthy brain development as well as our understanding of psychiatric and neurological conditions. Beyond individualized predictions, advantages of normative modelling include the transfer of predictions to unseen sites, data harmonization, and flexibility and adaptability to various brain imaging data types and distributions. However, the considerable power of these models requires careful handling and interpretation.

This course will educate attendees on the theoretical background and the fitting, use and interpretation of normative models. While the course will be taught with data provided, a strong focus will be on the transferability of skills to participants’ own data and research.

The overall objective is to provide participants with hands-on experience and tools to be able to transfer knowledge and use normative modelling on their own data.

Objectives

• Introduction to Normative Modeling: We aim to acquaint participants with the concept of normative modelling and showcase situations where its application is both appropriate and beneficial.
• Comprehensive Understanding of Methods: We will introduce the most common normative modelling methods, providing practical insights into their implementation. This encompasses a spectrum of approaches, ranging from purely statistical to Bayesian methods, with a focus on highlighting their respective strengths and potential challenges.
• Visualization and interpretation of Results: We will guide participants in understanding and interpreting the results of the models, emphasizing their relevance in subsequent analyses. Special attention will be given to creating visualizations and to illuminating potential limitations that participants should be aware of.

Advanced theory and applications of normative modelling:

• We will further illustrate advanced and applied uses of normative modelling. These will include:
  • Longitudinal models
  • Application to diverse brain imaging and non-imaging measures, such as
    • the functional connectome
    • executive functioning
    • and in relation with brain/body age
  • Advanced modelling approaches, such as using GAMLSS, Bayesian Linear Regression and Hierarchical Bayesian Regression models.
  • Site effect correction: Transfer and prediction of normative modeling results to new and unseen sites and populations

Curriculum

Overall time allocation: 4hrs

0. Settling in: 9am - 9.10am

1. Introduction to normative modelling and applying pre-trained normative models: 9.10am - 9.40am

• Duration: 30min;
• Presenter: Charlotte Fraza;
• Format: Jupyter notebooks in follow-along format;
• Programming language: Python;
• Libraries and Toolboxes: PCNToolkit and various standard Python libraries;
• Degree of interactivity: 50%.

2. Normative Models of Adolescent Executive Function Development Across Assessments and Datasets: 9.40am - 10.20am

• Duration: 40 min;
• Presenter: Brenden Tervo-Clemmens;
• Format: Talk and Live Demonstration in RMarkdown in follow-along format
• Programming language: R Language;
• Libraries and Toolboxes: various statistical R libraries,
• Degree of interactivity: 60%

3. Making longitudinal inferences using normative models: 10.20am - 10.50am

• Duration: 30 min;
• Presenter: Barbora Rehák Bučková;
• Format: Jupyter notebooks in follow-along format;
• Programming language: Python;
• Libraries and Toolboxes: PCNToolkit and various standard Python libraries;
• Degree of interactivity: 100%)

Break: 10.50am - 11.10am

• Duration: 20 min

4. Charting life-course functional connectome using normative modeling: 11.10am - 11.50am

• Duration: 40min;
• Presenter Lianglong Sun; Format:
• Talk and Live Demonstration in R and Markdown in follow-along format;
• Programming language: R Language;
• Libraries and Toolboxes: R package: gamlss;
• Degree of interactivity: 20%)

5. Methodological advances in normative modelling: Site effect correction: 11.50am - 12.20pm

• Duration: 30min;
• Presenter: Johanna M. M. Bayer;
• Format: Jupyter notebooks in follow-along format;
• Programming language: Python;
• Libraries and Toolboxes: PCNToolkit and various standard Python libraries;
• Degree of interactivity: 100% )

6. Characterizing brain-body relationships using normative modeling: 12.20pm - 1pm

• Duration: 40min; Presenter: Ye Ella Tian;
• Format: Interactive Presentation;
• Programming language, Toolboxes and Libraries: N/A:;
• Degree of interactivity: 20%)