Title
Multivariate Power Analysis for PRISME
Short description and the goals for the OHBM BrainHack
PRISME is an empirical power analysis toolbox for neuroimaging that estimates statistical power through subsampling of real data, with proven convergence to true power. Currently, PRISME supports univariate, network, and whole-brain inference methods. This project extends PRISME to support any multivariate statistical tests, adding a new analysis path to the core subsampling algorithm.
Link to the Project
https://github.com/neuroprismlab/PRISME-Brain-Power-Calculator
Image/Logo for the OHBM brainhack website
No response
Project lead
Fabricio Cravo, fabriciocravo, fabriciocravo
Main Hub
Bordeaux
Link to the Project pitch
No response
Other hubs covered by the leaders
Skills
Matlab or Python coding alone are enough, some background in multivariate statistical methods and mathematical analysis helps, but it is not required.
Recommended tutorials for new contributors
A quick glance at the original publication, to understand the original algorithm:
https://www.biorxiv.org/content/10.64898/2026.01.21.700878v1
And a glance at the toolbox documentation:
https://neuroprismlab.github.io/PRISME-Brain-Power-Calculator/
Good first issues
- Read the PRISME documentation and run the existing examples to understand the empirical subsampling pipeline
- Implement a parametric multivariate power calculator in Python or MATLAB with a known analytical power value, to serve as a validation baseline for the new pipeline
- Write unit tests for the multivariate path once the first test is implemented
Twitter summary
Extending PRISME, an empirical power analysis toolbox for neuroimaging, to support multivariate statistical tests. Join us at #OHBMBrainhack Bordeaux to help build and validate the new pipeline! 🧠📊 #OHBM2026
Short name for the Discord chat channel (~15 chars)
Multivariate Power Rangers
Please read and follow the OHBM Code of Conduct
Title
Multivariate Power Analysis for PRISME
Short description and the goals for the OHBM BrainHack
PRISME is an empirical power analysis toolbox for neuroimaging that estimates statistical power through subsampling of real data, with proven convergence to true power. Currently, PRISME supports univariate, network, and whole-brain inference methods. This project extends PRISME to support any multivariate statistical tests, adding a new analysis path to the core subsampling algorithm.
Link to the Project
https://github.com/neuroprismlab/PRISME-Brain-Power-Calculator
Image/Logo for the OHBM brainhack website
No response
Project lead
Fabricio Cravo, fabriciocravo, fabriciocravo
Main Hub
Bordeaux
Link to the Project pitch
No response
Other hubs covered by the leaders
Skills
Matlab or Python coding alone are enough, some background in multivariate statistical methods and mathematical analysis helps, but it is not required.
Recommended tutorials for new contributors
A quick glance at the original publication, to understand the original algorithm:
https://www.biorxiv.org/content/10.64898/2026.01.21.700878v1
And a glance at the toolbox documentation:
https://neuroprismlab.github.io/PRISME-Brain-Power-Calculator/
Good first issues
Twitter summary
Extending PRISME, an empirical power analysis toolbox for neuroimaging, to support multivariate statistical tests. Join us at #OHBMBrainhack Bordeaux to help build and validate the new pipeline! 🧠📊 #OHBM2026
Short name for the Discord chat channel (~15 chars)
Multivariate Power Rangers
Please read and follow the OHBM Code of Conduct