Postdoctoral Research Associate - AI for Medical Imaging & Biomedical Informatics

About the role

Dr. Pritam Mukherjee's lab is seeking a highly motivated Postdoctoral Research Fellow to develop advanced machine learning (ML), deep learning (DL), and foundational AI models for medical imaging, with a primary focus on radiology and a broader interest in biomedical informatics.

This role is well-suited to someone who thinks like a researcher first: someone who tracks the frontier of AI research (including large language models and multimodal foundation models) and translates emerging methods into novel approaches for imaging problems such as segmentation, quantification, detection, and image-based biomarker discovery across CT, MRI, and X-ray.

We welcome applicants from both PhD and MD (or MD/PhD) backgrounds.

This position sits within a well-resourced, data-rich research environment, with access to large, multi-institutional datasets and St. Jude Children's Hospital's excellent high-performance computing resources. Some datasets will already be curated, but others will require the lab to perform its own data curation and annotation, so experience with these tasks is a valuable asset.

This role offers strong support for career development: the lab is happy to support and encourage postdocs interested in pursuing independent funding, such as a K99/R00 or similar career development award, and there are opportunities to gain mentorship experience (e.g., with junior lab members) for those interested, though this is not a requirement of the role.

This is a 2–3 year appointment, with a flexible start date.

Key Responsibilities

• Design and implement 2D and 3D model architectures (CNNs, transformer-based, and foundational/vision-language models) for medical imaging problems
• Apply and adapt large language models (LLMs) to biomedical and imaging-adjacent problems, including training, fine-tuning, and inference/deployment of LLMs
• Stay current with the broader AI research literature (imaging, LLMs, multimodal foundation models) and bring new methods into ongoing research projects
• Build scalable pipelines for data preprocessing, model training, evaluation, and deployment
• Develop quantitative imaging methods (e.g., volumetrics, density measurements, biomarker extraction)
• Leverage curated, multi-institutional datasets, and contribute to data curation/annotation efforts where needed
• Collaborate with radiologists, clinicians, and engineering teams to ground research in what is actually useful and feasible in clinical practice
• Ensure reproducibility and traceability of experiments (data, model, and code versioning)
• Utilize modern AI-assisted development tools (e.g., LLM-based coding agents) to accelerate development and improve code quality
• Participate in team-based development practices (code reviews, Git, testing frameworks)
• Contribute to manuscripts, grant applications (including independent funding opportunities, if desired), and technical reporting
• Optional: gain mentorship experience by supporting junior lab members (PhD students, research assistants), for those interested

Required Qualifications

• PhD or MD (or MD/PhD) in Computer Science, Data Science, Biomedical Engineering, Medicine, or a related field
• Demonstrated experience with segmentation, detection, and/or quantitative imaging algorithms (2D and/or 3D)
• Strong proficiency in Python and deep learning frameworks (e.g., PyTorch, TensorFlow)
• Experience with modern architectures (U-Net variants, detection frameworks, transformers, or foundational models)
• Demonstrated, up-to-date knowledge of the AI research landscape, including recent developments in foundation models and/or LLMs
• Strong understanding of evaluation metrics relevant to imaging AI (Dice, IoU, ROC/AUC, sensitivity/specificity)
• Experience with version control and collaborative development (e.g., Git)

Additional information

Preferred Qualifications Candidates are not expected to have all of the following, but experience with one or more is a strong plus:

• Experience training, fine-tuning, and/or deploying LLMs (inference optimization, prompt/agent design, or related work)
• Experience with imaging modalities beyond radiology (e.g., microscopy, pathology, or other biomedical imaging)
• Experience or interest in working directly with clinicians, with some clinical knowledge or exposure that helps distinguish which research directions and model outputs are actually useful in practice
• Familiarity with DICOM and medical imaging workflows
• Broader background in biomedical informatics (e.g., multi-omics, EHR data, clinical NLP)
• Experience with large, multi-institutional datasets, including data curation and annotation
• Experience with cloud or high-performance computing environments
• Experience deploying models into research environments
• Exposure to translational AI development or regulatory-aware research (e.g., FDA pathways), though this is not the focus of the role

Academic and Career Development Opportunities

• Significant opportunities for first-authorship on high-impact manuscripts
• Active participation in multi-institutional research collaborations
• Dedicated support for pursuing independent funding (e.g., K99/R00 or similar career development awards) for those interested
• Optional mentorship opportunities (e.g., supervising or supporting PhD students or research assistants) for those interested in building mentoring experience
• Ability to build a strong academic portfolio, with the option to pursue translational/clinical impact where projects warrant it

Key Attributes

• Intellectually curious, research-driven, and genuinely engaged with the current AI literature

We are committed to a human-centered hiring experience. Technology may support portions of our process, but recruiting decisions involve human review and engagement. Learn more about our approach to AI.