top of page

RESEARCH: 

Biological Fluid Dynamics + Machine Learning 

(Giving a Stokes flow fluids demo at a Science Expo

for elementary school kids and their parents) 

Most of our research falls under the umbrella of mathematical biology, where we focus on applications from biomechanics, mathematical physiology, and numerical methods for PDE.

 

Lately, we've been creating and analyzing models of aquatic locomotion using machine learning techniques, while still developing popular software tools for the greater scientific community. 

​

Our lab collaborates with undergraduate and graduate students, postdocs, and faculty researchers across many disciplines, ranging from mathematics, engineering, pathology and laboratory medicine, biology, and physics. 

​

If you are interested in possibly working together on a project or chatting about research, please send me an email!

Using a variety of machine learning tools we are able to accelerate data collection and discovery to holistically explore the swimming behaviors and performance for a variety of animals. In particular, we test mechanical and evolutionary hypotheses while simultaneously investigating the diversity within biological systems. 

 

In the process we make extensive use of TCNJ's high-performance computing cluster, ELSA ("super computer")

NN_Cartoon.png

A very vanilla neural network architecture. We use a variety of different neural networks, including a variety of deep learning architectures, convolutional neural networks, and recurrent neural networks.

Some of the other ML tools and computational methods we regularly use include polynomial chaos expansions, supervised or unsupervised learning for either regression or classification, parameter estimation, and model learning.

 

Once we train, validate, and thoroughly test a surrogate model (whether neural network or polynomial chaos based), it opens the door for us to perform uncertainty quantification and global sensitivity analysis

We study a number of different swimming organisms including: jellyfish, polychaetes (marine worms), siphonophores, crustaceans (like krill or brine shrimp), anguilliform swimmers (like eels).

​

We also investigate many other biological fluid dynamics systems, ranging from designing and optimizing bioinspired pumps to exploring fluid flows around (or through) systems with complex morphologies, such as sea fans, starfish, or complex vascular networks like insect wings, to biooceanography systems, like copepod-marine snow interactions.

Overall, we are very interested in elucidating the underlying physics in living systems. 

FSI_CFD_Sims_2023.png

PREVIOUS PROJECTS **

Heroin/Opioid Epidemics

Opioid_Model_Schematic.png

Data-driven models describing the spread of opioids​ through various subgroups within a community. We are interested in community intervention strategies to most effectively thwart on-going escalation of the crisis. 

Tubular Heart Pumping

Comparing the mechanics of tubular pumping (impedance vs. peristaltic pumps) across various fluid regimes, including the underlying electrophysiology, in both vertebrate and invertebrate tubular hearts.

Ventricular Flows Across Complex Morphologies

Understanding how trabeculation affects cardiac flows embryonic zebrafish ventricles, in particular how it affects force (shear stress or pressure) distributions during heart morphogenesis.

** We are still interested in these applications, but have not actively worked on them in a bit.

For a list of our publications, books, or scientific computing notes, please click here:

bottom of page