Samuel W. Remedios · ML Researcher

Generative modeling and inverse problems for medical imaging and beyond

I pair data-driven generative priors with hard consistency to known physics, and make methods work when paired data is scarce, built on structure that transfers across scientific domains. Ph.D., Johns Hopkins.

See my work → Publications CV

A few things I've built, with code you can use

ECLARE

Published · J. Medical Imaging

Self-supervised through-plane MRI super-resolution from a scan's own slices. pip install eclare.

Paper Code

MedForj

In revision · Nature Sci. Reports

A large-scale 3D brain-MRI diffusion model built to be a reusable prior. Weights on Hugging Face.

Paper Code

PRISM

Published · MICCAI 2026

Contrast-invariant MRI slice-thickness estimation via spectral matching; differentiable PSF in PyTorch.

Paper Code

Always happy to talk generative modeling, inverse problems, and hard real-world ML. Reach me on LinkedIn or by email.