CMIC NACCs 🧠⌛️

[sophmrtn]

CMIC NACCs: Modelling brain heterogeneity in real-time 🧠⌛️

Leaders: Serena Verdi & Sophie Martin

📝 Description:

We can map brain variation at the individual level using neuroanatomical normative modelling. This project will attempt to speed up the process to give dementia patients a personalised score of brain heterogeneity in ‘real time’.

👉🏽 Requirements:

1. CS cluster access
2. A Google account
3. Some UNIX experience
4. Either Python or R experience

🗂 Resources we will provide:

• ~1600 T1w brain MRI scans from NACC
• Normative reference dataset of FreeSurfer cortical thickness and volume outputs from 58k scans
• Normative modelling script (Google colab)

☑️ Issues:

• [x] Generate brain thickness and volume metrics on NACC data using FastSurfer
• [x] Use NACC brain thickness and volume data within normative modelling pipeline to generate Z-scores
• [x] Map z-score data on brain surfaces

[sophmrtn]

Useful Links

• NACC UDS Variable Handbook
• NACC Imaging Data Variable Handbook
• How to run Fastsurfer guide
• Fastsurfer GitHub page
• Normative modelling application paper
• Normative modelling rationale paper
• Normative modelling Google Colab Notebook

[serenaverdi]

CMICHACKS slides.pptx