CAT12 Standalone Wrapper (Preprocessing + Longitudinal Statistics)

This repository provides a standalone, reproducible wrapper around CAT12 (SPM12) for:

• BIDS-compatible preprocessing (cross-sectional and longitudinal, auto-detected)

• Longitudinal group statistics (VBM + surface modalities) with screening and TFCE correction

It is designed for headless Linux servers and does not require a MATLAB license when using the bundled CAT12 standalone + MATLAB Runtime.

Getting started

• Installation
  • License and third-party software
  • System requirements
  • What gets installed (workspace-local)
  • Apple Silicon (ARM64) Setup
  • Install
  • Activate (optional)
  • Notes on reproducibility
  • Common installation issues
• Workflows (Preprocessing → Longitudinal Statistics)
  • Workflow A — Preprocessing (BIDS App style)
  • Workflow B — Longitudinal statistics
  • Recommended end-to-end pattern
• Usage — Preprocessing (cat12_prepro)
  • Synopsis
  • Processing stages (opt-in)
  • Selection options
  • Execution options
  • Examples
  • Help
• Usage — Longitudinal statistics (cat12_stats)
  • Two invocation modes
  • Required arguments
  • Analysis options
  • Design options
  • Screening options
  • TFCE options
  • Other options
  • Output layout
  • Examples
  • Post-Stats & Reporting
  • Help
  • Configuration defaults
  • Post-Stats Reporting
• Troubleshooting
  • Installation
  • Preprocessing
  • Statistics

Reference

• Quick Start Guide
• BIDS App Usage Guide
• CAT12 Standalone Commands Reference
• CAT12 BIDS Pipeline - Complete Summary

What you get

• Two entrypoints

  • ./cat12_prepro: preprocessing (BIDS App style)

  • ./cat12_stats: longitudinal group analysis

• Contained install under the repo directory:

  • external/ for CAT12 + MATLAB Runtime (+ local Deno)

  • .venv/ for Python dependencies

When to use which workflow

• Run preprocessing first to produce CAT12 derivatives.

• Then run statistics on the derivatives (typically vbm or thickness) and your participants.tsv.