Neonatal Normalization

A neonatal-to-adult normalization pipeline designed to minimize systematic biases in multi-step registration, enabling reliable cross-age comparisons in large-scale neuroimaging studies.

Background

In large-scale neuroimaging studies — especially those involving longitudinal analysis — a critical first step is to transform individual brain images into a common template space. This standardization enables meaningful group-level comparisons and statistical analyses.

For adult brains, this is typically achieved through one-step direct normalization to a standard template. This approach, however, does not generalize well to neonatal brains: due to substantial anatomical differences between neonatal and adult brains, direct normalization to an adult template often fails or yields suboptimal results.

A commonly adopted solution is a multi-step strategy: first normalize neonatal brains to age-specific templates (e.g., the UNC BCP 4D atlas), then progressively map them to older templates in a stepwise manner until reaching adult template space. This has been shown to improve registration performance.

However, the multi-step approach may introduce systematic biases. If not properly addressed, these biases can propagate into downstream analyses and be mistakenly interpreted as biological signals.

Objective

Develop a neonatal normalization pipeline that:

• minimizes systematic biases introduced by multi-step registration, and
• achieves consistency with adult normalization,

thereby enabling more accurate and reliable cross-age comparisons.

Pipeline Overview

1. Skull stripping — extract brain tissue using iBEAT 2.0.
2. Age-specific normalization — register each subject to the corresponding age-specific template (UNC BCP 4D atlas).
3. Visual quality check — inspect skull-stripping and registration results.
4. Stepwise mapping to adult space — progressively map to older templates up to the adult template.

Repository Structure

NeonatalNormalization/
├── Notebook/
│   └── SickKids_tutorial.ipynb     # Walkthrough / tutorial
├── Script/
│   ├── pre01_iBEAT/                # iBEAT 2.0 preprocessing (skull stripping)
│   └── tpl_xfm_build.py            # Template transform construction
└── templates/                      # Age-specific & adult templates (BCP atlas)