Fetpype: An Open-Source Pipeline for Reproducible Fetal Brain MRI Analysis

Fetal brain magnetic resonance imaging (MRI) is crucial for assessing neurodevelopment in utero. However, fetal MRI analysis remains technically challenging due to fetal motion, low signal-to-noise ratio, and the need for complex multi-step processing pipelines. These pipelines typically include motion correction, super-resolution reconstruction, tissue segmentation, and cortical surface extraction. While specialized tools exist for each individual processing step, integrating them into a robust, reproducible, and user-friendly end-to-end workflow remains difficult. This fragmentation limits reproducibility across studies and hinders the adoption of advanced fetal neuroimaging methods in both research and clinical contexts. Fetpype addresses this gap by providing a standardized, modular, and reproducible framework for fetal brain MRI preprocessing and analysis, enabling researchers to process raw T2-weighted acquisitions through to derived volumetric and surface-based outputs within a unified workflow. Fetpype is publicly available on GitHub at https://github.com/fetpype/fetpype.

💡 Research Summary

The manuscript introduces Fetpype, an open‑source, Python‑based framework that unifies the entire fetal brain magnetic resonance imaging (MRI) processing chain—from raw T2‑weighted stacks to volumetric and surface‑based outputs—into a single, reproducible workflow. The authors begin by outlining the unique challenges of fetal MRI: severe intra‑subject motion, low signal‑to‑noise ratio, and the necessity of multiple processing stages (motion correction, super‑resolution reconstruction, tissue segmentation, cortical surface extraction). Existing tools typically address only one of these stages and require custom scripting to glue them together, which hampers reproducibility and limits multi‑site collaborations.

Fetpype tackles these issues through four core design principles:

1. Data Standardization – Input data must follow the Brain Imaging Data Structure (BIDS) specification, ensuring consistent metadata, naming conventions, and directory hierarchy across sites.

2. Containerization – Each processing component (brain extraction with Fetal‑BET, non‑local means denoising, N4 bias‑field correction, super‑resolution reconstruction, segmentation, surface extraction) is encapsulated in Docker or Singularity containers. This isolates software dependencies, eliminates “works on my machine” problems, and simplifies installation on both workstations and high‑performance clusters.

3. Workflow Orchestration – The Nipype engine constructs a directed acyclic graph linking the containers, providing automatic data caching, parallel execution, and provenance tracking. Nipype’s modular interfaces also allow seamless substitution of alternative tools.

4. Flexible Configuration – Hydra together with human‑readable YAML files lets users select algorithms (e.g., choose among NeSVOR, SVR‑TK, or NiftyMIC for super‑resolution), adjust hyper‑parameters, and enable/disable modules without touching source code.

The pipeline currently integrates:

• Pre‑processing – Fetal‑BET for brain masking, adaptive non‑local means denoising, and N4ITK bias correction.
• Super‑Resolution Reconstruction – Three widely used approaches (NeSVOR, SVR‑TK, NiftyMIC) are provided, giving users flexibility to match scanner characteristics and acquisition protocols.
• Segmentation – Both the BOUNTI volumetry/parcellation suite and the Developing Human Connectome Project (DHCP) neonatal pipeline are wrapped, delivering tissue class maps and quantitative volumes.
• Cortical Surface Extraction – A custom implementation based on prior work (Bazin & Pham, 2005/2007; Ma et al., 2022) generates 3D cortical meshes suitable for downstream morphometric analyses.

The authors validated Fetpype on heterogeneous datasets collected across France, Spain, and Switzerland, encompassing multiple scanner models and gestational age ranges. The pipeline has already been employed in a large consortium‑wide study (see reference