Software Developer - Cloud-Based MRI Reconstruction Pipeline

The Lingala Lab in Biomedical Engineering and Radiology, in collaboration with Champaign Imaging, is seeking a motivated software developer (temporary professional staff role or MS student in Computer Science / Computer Engineering / Electrical & Computer Engineering to help build a vendor-neutral, cloud-capable MRI image reconstruction pipeline. This project is funded through NIH. This is a hands-on systems/software role for someone who wants real exposure to medical imaging infrastructure without needing an MRI physics background.

The core idea: our lab acquires raw Fourier-space MRI data on GE scanners at the Magnetic Resonance Research Facility, and our current reconstructions are typically tied to vendor-specific on-scanner software or custom off-line in-lab software. We want to decouple reconstruction from the scanner by containerizing our pipeline with Docker, running it through the open-source reconstruction engines, and deploying it both locally and to the cloud using Microsoft's Tyger framework (https://microsoft.github.io/tyger/ (https://microsoft.github.io/tyger/)). This mirrors a recently published approach where a research group containerized their reconstruction pipeline with Gadgetron, pushed it to a registry, and used Tyger to run it as a cloud job - while carefully benchmarking compute time vs. data-transfer time between local and cloud execution.

* Get up to speed on Gadgetron (the open-source, vendor-neutral MRI reconstruction engine we're standardizing on) and its "gadget" pipeline architecture.

* Package Gadgetron (plus our custom reconstruction code) into a Docker container that runs identically locally and in the cloud.

* Push containers to a registry and set up Tyger to pull and run them as cloud jobs - starting from Tyger's own demos to learn the submit run retrieve workflow, then moving to our real scanner datasets.

* Benchmark and quantify the pipeline: separate compute time (actual reconstruction) from data transfer time (raw data up, images back down), and compare local vs. cloud execution.

* Work directly with our team to eventually route custom, lab-developed reconstruction algorithms through this pipeline instead of as one-off standalone scripts.

* Document the setup so it's reproducible across our GE MRI scanners at the MRRF.

* This positioncollaborates with Dr. Sajan Lingala (BME, Radiology), Dr. Jim Holmes (Radiology, BME), Collaborators at Champaign Imaging, and other student/post-docs within MRRF.

Required:

* Strong programming background (Python and/or C++); comfortable reading and modifying existing open-source codebases.

* Hands-on experience with Docker (building images, writing Dockerfiles, running containers).

* Familiarity with cloud computing concepts (job submission, registries, YAML-based configuration) - direct Azure/AWS/GCP experience is a plus but not required.

* Experience with Linux server administration and continuous integration/continuous deployment pipelines.

* Comfortable working semi-independently, reading technical documentation (including reading actual source code / GitHub repos), and iterating.

Desirable:

* Coursework or interest in signal/image processing.

* Prior exposure to MRI, medical imaging

No MRI or medical imaging background is required - we will train you on the domain-specific parts. We are primarily looking for strong software engineering fundamentals and genuine interest in systems/infrastructure work.

* Temporary part-time professional position

* Duration: 1 year, with possible 1-year extension.

* Salary: 20$/hr; 20hours a week;

* Location: University of Iowa MRRF / BME, Iowa City, IA. Primarily in-person (with on-scanner work), with some remote-friendly development work possible.