Biomarkers for Neurodegenerative Disorders

This repository hosts the course materials for the biostatistics module of the Biomarkers for Neurodegenerative Disorders course in Gothenburg, Sweden, May 20 to 24 2024.

Statistical and visualization techniques will be demonstrated using R applied to data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). If you would like to execute the R code on your own machine, you should apply for access to ADNI and download and install the ADNIMERGE R package.

RStudio is also highly recommended.

The biostatistics module topics include:

Biostatistics for fluid and imaging biomarkers (hour 1, Mike Donohue)

  • Batch Effects
  • Experimental Design (Sample Randomization)
  • Statistical Models for Assay Calibration/Quantification
  • Classification (Supervised Learning)
    • Logistic Regression
    • Binary Trees
    • Random Forest
  • Mixture Modeling (Unsupervised Learning)
    • Univariate
    • Bivariate
    • With covariates (Mixture of Experts)
  • Reference Regions
  • Centiloids
  • Standardization using the Empirical Cumulative Distribution Function (ECDF)

Longitudinal Data (hour 2, Lars Racket)

  • Mixed effect models
  • Mean & Variance Structure
  • Mixed Model for Repeated Measures (MMRM)
  • Constrained Longitudinal Analysis (cLDA)
  • Model selection strategies
  • Nested random effects
  • Missing Data, MAR, MNAR
  • Multiple Imputation & Delta Method