Postdoctoral Fellow-MSH-32030-004
About The Position
We are seeking a highly motivated Postdoctoral Research Fellow to join the Medical Intelligence lab and contribute to research at the intersection of medical imaging analysis and machine learning. The successful candidate will design, train, and validate deep learning models on clinical imaging data, publish in leading venues, and collaborate with clinicians, data scientists, and engineers. This role is well suited to someone who wants to translate methodological advances into tools with real clinical impact.
Project Description
This position centers on building large-scale foundation models for medical imaging and translating them into clinically meaningful applications. The work spans three connected threads:
- Large-scale vision-language model development. Design and train large multimodal vision-language models that jointly reason over medical images and associated text (reports, clinical notes, structured data), with an emphasis on scalable pretraining, efficient fine-tuning, and robust evaluation.
- Retinal imaging foundation model. Build and adapt a foundation model for retinal imaging (fundus photography and OCT) that can be pretrained on large image collections and transferred efficiently to a range of downstream tasks with limited labeled data.
- Downstream clinical applications. Apply and adapt these models to real-world clinical problems, including systemic and hematologic conditions such as multiple myeloma and sickle cell disease, where retinal and multimodal biomarkers may support early detection, risk stratification, and disease monitoring.
The successful candidate will help move the group's work from general-purpose model development toward validated, clinically relevant tools, working closely with clinical collaborators throughout.
Responsibilities
- Develop, train, and evaluate deep learning models for medical image analysis (classification, segmentation, detection, and related tasks).
- Build reproducible experimental pipelines in PyTorch, including data preprocessing, model training, and rigorous validation.
- Curate, clean, and manage imaging datasets while adhering to data governance, privacy, and ethics requirements.
- Design and run experiments, analyze results, and iterate on model architectures and training strategies.
- Write and publish first-author papers in peer-reviewed journals and top-tier conferences.
- Contribute to grant proposals, progress reports, and presentations to internal and external stakeholders.
- Collaborate with clinical partners to define problems, interpret results, and ensure clinical relevance.
Required Qualifications
- PhD in Computer Science, Biomedical Engineering, Electrical Engineering, Applied Mathematics, Medical Physics, or a closely related field (completed, or defended before the start date).
- Demonstrated experience in medical imaging analysis (e.g., MRI, CT, X-ray, ultrasound, OCT, or fundus imaging).
- Strong proficiency in PyTorch and modern deep learning workflows.
- Solid programming skills in Python and familiarity with the scientific computing stack (NumPy, pandas, scikit-learn, etc.).
- Track record of peer-reviewed publications appropriate to career stage.
- Strong analytical, written, and verbal communication skills, and the ability to work both independently and collaboratively.
Preferred Qualifications
- Experience working with electronic health records (EHR) and integrating structured clinical data with imaging (multimodal modeling).
- Experience with retinal imaging (fundus photography, OCT) and related tasks such as diabetic retinopathy or glaucoma analysis.
- Familiarity with medical data standards and privacy frameworks (e.g., DICOM, HL7/FHIR, HIPAA/GDPR).
- Experience with segmentation frameworks, self-/semi-supervised learning, foundation models, or model interpretability.
- Experience with version control (Git), containerization (Docker), and HPC or cloud GPU environments.
What We Offer
- A collaborative, interdisciplinary research environment with access to clinical data and domain experts.
- Access to GPU clusters / compute resources / imaging datasets.
- Support for conference travel, publication, and professional development.
- Dedicated mentorship and a clear commitment to your growth as an independent researcher. We will support you in developing your research agenda, publishing high-impact work, building collaborations, and preparing for the next step in your career, whether in academia or industry.
- Mount Sinai standard benefits package, relocation support.
How To Apply
Please submit the following to xueyan.mei@mssm.edu with the subject line "Postdoc — Medical Imaging & Deep Learning":
- Cover letter describing your research interests and fit for the role.
- Curriculum vitae, including a full publication list.
- Names and contact details of 2–3 references.
- (Optional) Links to representative code, projects, or a Google Scholar profile.