Souvik Roy
University of Texas at Arlington, TX·Mathematics & Statistics
Light and sound can work together to image tumors. In quantitative photoacoustic tomography (QPAT), short pulses of near-infrared light are absorbed by tissue.
- Fields
Mathematics & Statistics
Biomedical Engineering
Engineering
- Topics
quantitative photoacoustic tomography
photoacoustic imaging
inverse problems
- Grants
- 1 known grant, through Aug 2027
What they work on
- Develop a gradient-free Nash game computational scheme that fills in missing photoacoustic measurements and estimates the unknown sound speed and optical energy density in tissue.
- Build a gradient-free optimization method to reconstruct light absorption and scattering profiles in QPAT with high contrast and resolution. The method avoids costly derivative calculations.
- Use statistical sensitivity analysis, which measures how outputs depend on uncertain inputs, to stabilize and calibrate the Nash algorithm so that QPAT reconstructions are reliable.
- Validate the computational reconstruction framework using real photoacoustic imaging data from mouse specimens and compare its quality and cost against existing methods.
More about the research
The tissue heats slightly and gives off ultrasound waves, and those waves can be measured and turned back into a map of how the tissue absorbs and scatters light. Such maps can carry useful information about cancerous tissue for diagnosis, staging and treatment planning. Turning the sound measurements back into an image is a hard math problem, called an inverse problem. In practice, scanners collect incomplete data, and the speed of sound in tissue is not known exactly. Both limits make current reconstructions less accurate.
Why it matters
Most cancers show no early symptoms, so many are diagnosed late. Clearer quantitative images of tumor tissue from photoacoustic imaging could help clinicians detect cancer and plan its treatment. Reconstruction methods that are computationally cheap and robust to incomplete data and uncertain sound speed could make this imaging approach more practical. More broadly, the methods could improve other limited-data imaging problems.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Mark Anastasio
Washington University in St Louis, MO·Biomedical Engineering
Develop multi-physics simulation tools that model light transport and sound detection to generate realistic 3D breast optoacoustic tomography (OAT) measurement data from numerical breast phantoms.
Stephen Becker
University of Colorado Boulder, CO·Mathematics & Statistics
Develop gradient-free optimization algorithms for objective functions whose matrix-valued gradients have low intrinsic dimensionality, such as sparsity, low rank, or sparsity-plus-low-rank, to break worst-case efficiency barriers.
Yuhua Zhu
University of California-Los Angeles, CA·Mathematics & Statistics
Develop a consensus-based interacting particle method, where many searchers share information to agree on the best point, for nonconvex optimization problems with equality constraints.
The school
University of Texas at Arlington: 11 labs on PI Pages, $154M a year on research, 177 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.