<aside> 👩🏻💻
Author: Zihan Ning, PhD
First released: 26-Aug-2025
Last updated: 26-Aug-2025
</aside>
Gadgetron is an open-source medical image reconstruction platform that allows researchers and developers to implement custom MR reconstructions within the clinical MR imaging workflow. Originally designed to support real-time and high-performance MRI processing, Gadgetron enables flexible reconstruction pipelines using vendor-neutral formats (e.g., ISMRMRD) and high-level languages like MATLAB and Python. Importantly, Gadgetron supports inline deployment—meaning that reconstructed images can be returned directly to the MR scanner console and then to PACS, enabling integration with routine clinical review and reporting.
Despite these strengths, converting a well-developed offline reconstruction script into an inline implementation remains a non-trivial task. This is where the proposed Gadgetron Parallel Reconstruction Framework comes in.
Built on top of the Gadgetron platform, this framework is designed to bridge the gap between offline prototyping and inline deployment, especially for users who already have established reconstruction code taken Siemens format raw (Twix) as the input. It provides:
The framework has been validated in both research and clinical settings, including routine sequences such as SENSE, AlignSENSE, and radial NUFFT reconstructions. It supports both single-sequence and multi-sequence protocols and has been stress-tested in large-scale studies like TwinsUK.
This user manual is structured to support both new users and experienced developers:
Offline Test
Learn how to simulate an inline reconstruction workflow on your local PC using exported Siemens .dat
files. Includes both Debug and Auto Modes to help with prototyping and automation.
✅ [Quick try-out] section gives detailed instructions to run the demos provided by the git repository.
Inline Deployment
Follow step-by-step instructions to configure Gadgetron on the scanner console and external server, including ICE integration and TCP/IP setup.
Custom Development Guide
Learn how to adapt your own reconstruction scripts for inline use with minimal code changes, and how to expand the input converter when required fields are missing.
Whether you're aiming to test reconstructions offline, deploy a custom pipeline inline, or translate an entire suite of tools for clinical integration, this manual is your guide. It is written to be practical, hands-on, and aligned with the open philosophy of the Gadgetron project.
Let’s get started 🥳 🐾