Postdoctoral Research Associate — Computational Vascular Geometry & Quantitative Aortic Imaging

The Advanced Aortic Imaging for Clinical and Translational Research (AAI-CTR) lab at UW–Madison is hiring a Postdoctoral Research Associate to develop next-generation computational methods for quantitative 3D analysis of aortic anatomy from serial CT and MRI. If you want your algorithms to change how patients are monitored and treated, not just generate papers (although there is plenty of that too), this is the role.

Why This Lab

Our group has built several image analysis techniques that have fundamentally changed how clinicians assess aortic disease. We develop spatially resolved, 3D imaging biomarkers that capture patterns of anatomical change invisible to conventional measurement, and we are actively translating these methods from research tools to clinical products. Our work sits at the intersection of computational geometry, shape analysis, and clinical decision-making, with efforts aimed at brining our discoveries to patients with aortic disease. Former trainees won a variety of prestigious awards, landed competitive industry positions, co-developed new IP, and gone on to train at other top-tier clinical programs. This is a lab where technical innovation and clinical impact are not separate goals.

What You’ll Do

Geometric Methods Development

● Develop and validate methods for robust 3D surface parameterization, anatomical correspondence, and spatially resolved geometric feature extraction from vascular imaging data.

● Design approaches for establishing reliable correspondence between serial imaging studies of the same patient, with emphasis on separating true anatomical change from positional and acquisition-related variation.

● Build computational pipelines for longitudinal shape comparison and quantitative morphometric analysis of the aorta using large, multi-center datasets.

● Benchmark novel methods against existing clinical reference standards; lead external validation and reproducibility testing.

Clinical Translation

● Collaborate directly with cardiovascular surgeons, cardiologists, and radiologists to define clinically meaningful performance targets for geometric biomarkers.

● Advance investigational tools toward real-world use: workflow integration, reliability testing, and documentation aligned with regulatory and commercial pathways.

● Apply novel analysis methods to large patient populations (UW and multi-center cohorts) to generate evidence that supports clinical adoption.

Data & Project Leadership

● Curate, organize, and manage large clinical imaging datasets and associated outcomes data.

● Mentor and collaborate with trainees across clinical and engineering backgrounds.

● Lead manuscript preparation, conference abstracts, and presentations; build your independent publication record.

Required Qualifications

● Ph.D. (or equivalent) in Computer Science, Biomedical Engineering, Applied Mathematics, Electrical/Computer Engineering, Medical Physics, or a closely related quantitative field.

● Strong foundation in computational geometry, differential geometry, or geometric modeling, with hands-on experience applying these methods to real data.

● Experience with 3D surface or mesh processing: parameterization, correspondence, curvature computation, shape descriptors, or related techniques.

● Proficiency in Python; comfort working with scientific computing libraries and building reproducible analysis pipelines.

● Experience with medical image analysis on CT and/or MRI, or with 3D geometric data from a related domain (biomechanics, computer graphics, computational anatomy).

● Track record of scientific productivity (peer-reviewed publications, conference papers) and strong written/oral communication.

● Genuine interest in clinical translation, meaning building tools clinicians will actually use, not just publishing methods papers.

Preferred Qualifications

● Experience with statistical shape modeling, shape analysis, or atlas-based methods.

● Background in robust alignment or pose estimation, particularly methods that separate rigid pose from intrinsic shape change.

● Familiarity with implicit surface representations (signed distance functions, level sets) or related continuous geometric representations.

● Prior work with vascular or cardiovascular imaging (aorta, aneurysm, cardiac CT/MR).

● Knowledge of image registration concepts (rigid, affine, deformable) and an understanding of their strengths and limitations.

● Experience with deep learning for medical image segmentation or geometric deep learning; familiarity with PyTorch or TensorFlow.

● Comfort working in multi-disciplinary teams and collaborating closely with clinicians.

● Creative, independent problem-solving mindset: able to prototype, test, and iterate quickly.

Commitment & Environment

We are looking for someone who can commit to at least 18–24 months. Meaningful translational research takes time: developing a method, validating it rigorously, and publishing the results is not a short-term endeavor. In return, you will work in a highly collaborative environment spanning Radiology, Surgery, Cardiology, and Medical Physics at UW–Madison, with direct access to large clinical datasets and a PI who is invested in your long-term career trajectory.

Work Authorization

Candidates who already hold valid U.S. work authorization are preferred, including F-1 OPT and STEM OPT. We are open to sponsoring H-1B visas for exceptional candidates, though candidates who do not require immediate employer-sponsored work authorization will have an advantage given current processing timelines.