Research
Research & Publications
Scientific work connecting medical computer vision, probabilistic inference, signal reconstruction, and mathematical optimization.
Research areas
- Medical computer vision
- Learning from biomedical images.
- Diffusion MRI
- Inference of tissue and fiber structure.
- Bayesian deep learning
- Probabilistic modeling and uncertainty.
- Compressed sensing
- Reconstruction from incomplete measurements.
- Statistical signal processing
- Estimation and inference from noisy signals.
- Power systems & scientific optimization
- Mathematical modeling of constrained systems.
Research themes
Harvard University · 2021 – 2022
Bayesian deep learning for diffusion MRI
Probabilistic, uncertainty-aware inference of brain fiber pathways from diffusion MRI.
- Bayesian deep learning framework for brain fiber pathway inference
- CNN voxel classification resolving multi-fiber crossings in noisy data
- Transfer learning for likelihood-free posterior estimation
Kansas State University · 2016 – 2020
Dynamic compressed sensing
Reconstructing signals and images from noisy, delayed, and incomplete measurements, with applications to accelerated medical imaging.
- Two dynamic compressed sensing reconstruction algorithms
- Error quantification and stability analysis via statistical signal processing and control theory
Kansas State University · 2016 – 2020
Estimation for smart power grids
Sparse reconstruction and optimization for observability in data-scarce distribution grids.
- Compressed sensing with estimated power flow for three-phase unbalanced grids
- Mixed-integer nonlinear framework for joint state estimation and topology identification