Software Developer

GAP Solutions Inc; Bethesda (MD); On-site; USD 110,000 - 180,000; Full time

Job summary

NHLBI DIR is seeking a software developer to design and maintain GPU-accelerated MRI reconstruction pipelines and CT interventional workflows. You will work on Gadgetron-based C++/CUDA modules and Python/PyTorch replacements, enabling research and clinical imaging.

You will build scanner-to-cloud integrations, data infrastructure for AI training, and data management for multi-site collaboration, reporting to NHLBI staff scientists.

Qualifications

• BS or higher in CS, engineering, physics, or related field.
• Proficiency in Python (NumPy, PyTorch, SciPy) and C++ (C++17+).
• GPU programming experience (CUDA, PyTorch GPU ops, or Triton).
• Familiarity with Linux (Ubuntu), Docker, and containerized workflows.
• Git/GitHub, CI/CD pipelines, and automated testing frameworks.
• Strong software engineering practices: code review, documentation, modular design, and test-driven development.

Responsibilities

• Develop and maintain MRI reconstruction modules in Python and C++ using the Gadgetron frameworks.
• Support scanner-to-compute integration bridging scanners with Azure-based and on-premises compute via the MRD/Gadgetron protocol.
• Build and maintain data ingestion pipelines, dataset management, and automated processing workflows for AI model training.
• Support software development for imaging platforms including near-real-time data streaming.
• Write software tools including solvers; port prototype code into deployment-ready software.
• Develop, test, and deploy software for clinical MRI and CT pipelines.
• Commit source code to designated Government repositories with documentation and unit tests.
• Participate in code reviews, sprint planning, and technical design discussions.
• Provide biweekly technical progress reports and deliverables documentation.
• Deliver Docker images and deployment artifacts for reconstruction pipelines as needed.

Skills

Python

C++

CUDA

PyTorch

Gadgetron

ISMRMRD/MRD

Linux (Ubuntu)

Docker

Git/GitHub

CI/CD pipelines

Education

B.S. or higher in CS/Engineering/Physics

Tools

Azure

DICOM

Siemens ICE/IDEA

Job description

Position Objective: The National Heart, Lung, and Blood Institute (NHLBI), Division of Intramural Research (DIR), conducts advanced medical imaging research spanning MRI and CT of the heart and lungs. Several research groups are focused on image acquisition, reconstruction, post-processing, and clinical application of X-ray, CT, and MRI images. The medical imaging programs within NHLBI DIR require software engineering and algorithm development infrastructure that supports these clinical and research imaging programs.

The NHLBI, Laboratory of Imaging Technology seeks a full-time software developer for cardiothoracic medical image reconstruction. The developer will build and maintain GPU-accelerated MRI reconstruction software, scanner-to-cloud integration systems, data infrastructure for AI model training, and CT reconstruction for interventional guidance. These systems serve both research teams within NHLBI and collaborators across the National Institutes of Health.

The candidate will be required to maintain and develop multiple software systems critical to NHLBI’s imaging mission, including image reconstruction pipelines, AI-enhanced image reconstruction methods, scanner-to-cloud integration systems, and data management infrastructure for AI model training. These systems serve both the research teams within NHLBI DIR and collaborators across the National Institutes of Health and partnering institutions.

The technical services required from the Software Developer candidate will support the design, implementation, testing, and maintenance of software systems across the clinical imaging pipeline work areas, with a specific focus on MRI reconstruction and interventional cardiovascular CT. The candidate will build and maintain clinical imaging pipelines that enable cardiovascular and pulmonary imaging research. The developer will focus on two primary areas:

• MRI Reconstruction Pipelines: Develop and maintain GPU-accelerated MRI reconstruction software (Gadgetron-based C++/CUDA and next-generation Python/PyTorch platform), scanner-to-cloud integration for clinical workflows, and supporting data infrastructure for AI model training.

• Interventional Cardiovascular CT: Develop data-streaming infrastructure between CT scanner and custom reconstruction server.

The objective is to provide software development services that directly support NHLBI’s ability to acquire, reconstruct, and analyze cardiac and pulmonary imaging data in both research and clinical settings.

The candidate shall perform software development tasks across the following work areas. All work shall be performed under the technical direction of NHLBI Staff Scientists.

MRI Reconstruction Pipeline (Gadgetron)

The candidate shall support the development and maintenance of NHLBI’s MRI reconstruction software, including the existing Gadgetron-based C++/CUDA pipeline and its next-generation Python/PyTorch replacement (Puffyn). Tasks include:

• Develop and maintain reconstruction modules (gadgets) in Python and C++ for cardiac and pulmonary MRI applications.

• Write unit tests, integration tests, and GPU-tier tests; maintain CI/CD pipelines (GitHub Actions, Docker, Azure Container Registry).

• Write software tools including solvers.

• Port prototype code into deployment-ready software.

• Maintain and test docker containers following system updates.

Scanner-to-Cloud Integration (Ice-Gadgetron/Tyger)

The candidate shall support the development of Ice-Tyger, and Ice-Gadgetron, systems that bridges on-scanner MRI reconstruction (Siemens ICE environment) with Azure-based cloud compute and on-prem compute via the MRD/Gadgetron protocol. Tasks include:

• Develop and maintain ICE functor libraries in C++ for the Siemens IDEA/ICE development environment.

• Test end-to-end data pathways from scanner acquisition through cloud reconstruction and image return.

Data Infrastructure and AI Training Pipelines

The candidate shall support the establishment and maintenance of data management systems that enable AI model development for medical imaging. Tasks include:

• Build and maintain data ingestion pipelines for DICOM and raw MRI/CT data

• Develop and manage datasets.

• Build automated processing pipelines for multi-stage workflows: reconstruction, analysis, and report generation.

• Support secure data sharing with external collaborators, including generation of read-only access credentials and transfer monitoring.

CT Reconstruction

The candidate shall support software development for a next-generation interventional cardiovascular CT platform. Tasks include:

• Support real-time or near-real-time data streaming.

• Collaborate with clinical and engineering staff on system integration and testing.

Duties and Responsibilities

• Develop and maintain MRI reconstruction modules in Python and C++ using the Gadgetron frameworks

• Support scanner-to-compute integration bridging scanners with Azure-based and on-premises compute via the MRD/Gadgetron protocol

• Build and maintain data ingestion pipelines, dataset management, and automated processing workflows for AI model training

• Support software development for imaging platforms including near-real-time data streaming

• Write software tools including solvers; port prototype code into deployment-ready software

• Develop, test, and deploy software for clinical MRI and CT pipelines.

• Commit source code to designated Government repositories (GitHub) with appropriate documentation, inline comments, and unit tests accompanying each submission.

• Participate in code reviews, sprint planning, and technical design discussions as directed.

• Provide biweekly technical progress reports summarizing completed tasks, issues encountered, and planned work for the upcoming period.

• Produce software documentation including architecture documents, API references, and user guides with each major release or as directed.

• Deliver test reports and CI/CD pipeline results with each code submission.

• Build and deliver Docker images and deployment artifacts for reconstruction pipelines as needed.

• Participate in regular team meetings and project planning sessions as directed.

• Deliverables:

• Source code committed to designated repositories with documentation and tests on ongoing basis

• Technical progress reports summarizing completed tasks, issues, and planned work on bi-weekly basis

• Software documentation (architecture documents, API references, user guides) with each major release

• Test reports and CI/CD pipeline results with each code submission

• Docker images and deployment artifacts for reconstruction pipelines as needed

Qualifications

Basic Qualifications

• B.S. or higher degree in computer science, engineering, physics, or a related discipline

• Proficiency in Python (NumPy, PyTorch, SciPy) and C++ (C++17 or later)

• Experience with GPU programming (CUDA, PyTorch GPU operations, or Triton).

• Familiarity with Linux system administration (Ubuntu), Docker, and containerized workflows.

• Experience with version control (Git/GitHub), CI/CD pipelines, and automated testing frameworks.

• Strong software engineering practices: code review, documentation, modular design, and test-driven development.

Preferred Qualifications

• Experience with MRI or medical imaging reconstruction software (Gadgetron, BART, SigPy, or similar).

• Familiarity with ISMRMRD/MRD data formats and the DICOM standard.

• Experience with Azure cloud services (Blob Storage, Container Instances, or equivalent).

• Familiarity with Siemens MRI development environments (ICE/IDEA) or other vendor SDKs.

This job description is not designed to cover or contain a comprehensive listing of activities, duties or responsibilities that are required by this position.

To perform this job successfully, an individual must be able to perform each essential duty satisfactorily. The requirements listed above are representative of the knowledge, skill, and/or ability required. Reasonable accommodations may be made to enable individuals with disabilities to perform the essential functions.

Qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability, protected veteran status or other characteristics protected by law.