Direct Analysis in Real Time Mass Spectroscopy (DART+ MS) is a chemical identification tool that uses a superheated gas stream to ionize chemical samples, producing a distinct chemical signal that can be used to identify the composition of an unknown sample. DART+ MS is used reliably in fields like forensics, food safety, pharmaceuticals, and more recently, environmental protection. At the Wasser Research Lab, at the Center for Environmental Forensic Science, we work to protect endangered species such as African Elephants. Using Direct Analysis in Real Time Mass Spectroscopy, we seek to find if elephant ivory from different regions in Africa has distinct chemical signatures, allowing us to geolocalize ivory samples based on their DART+ MS signatures. Current methods of elephant geolocation include genetic testing, but results can often be ambiguous; By using this completely different, complementary approach, we could improve our estimates of these inconclusive tests. If there is a chemical difference in the ivory of Elephants from the Savannah and Forest regions of Africa, then we can trace the origins of ivory obtained from illegal seizures, aiding in the conservation efforts of African elephants. Chemical distinctions aside, we also hope to answer questions about the effects of certain chemical preservatives on ivory samples and whether the DART+ MS signal varies along the length of the cut of the tusk, establishing best practices for sampling. Ultimately, our goal is to determine if DART+ MS proves to be a reliable and quick method of identifying elephant ivory for conservation efforts. By bridging cutting-edge technology with conservation science, we hope this research will make a significant impact on efforts to combat the illegal ivory trade and wildlife crime. 

Pulmonary Arterial Hypertension (PAH) is a deadly vascular disease, affecting the blood vessels of the lungs, with no existing cure. PAH is characterized by pulmonary arterial smooth muscle cell (PASMC) hypertrophy and hyperplasia, which increases resistance to blood flow within the pulmonary arteries, leading to rapid symptom progression and eventual death from right heart failure. My mentor and I hypothesize that defects in PASMC differentiation and alignment may contribute to PAH. To test whether alignment and phenotypic responses differ in patients with PAH, we designed a micropatterned collagen scaffold atop a glass coverslip. Explanted PASMCs from patients with PAH or failed donors (controls) were cultured on alternating 10-µm wide x 10-µm deep microchannels or unpatterned constructs and alignment, protein expression, and cellular morphology were compared across conditions. I evaluated 3 PAH and 3 control subjects and have collected preliminary data for each condition (control versus PAH), with three technical replicates each. Through these preliminary studies, I have demonstrated success of my model with consistent alignment observed on patterned substrates. Excitingly, PASMCs from patients with PAH expressed significantly decreased levels of the contractile protein, Calponin, when compared with control cells, including after responding to cues that promote alignment and contractility. This suggests that PAH PASMCs remain in an inappropriately synthetic or proliferative state. Moving forward, I plan to evaluate additional micropatterns by varying dimensions of rectangular and sine waves designs using an ablation protocol with a 2-photon microscope laser. Subsequent evaluation will include immunofluorescent staining of contractile and other SMC markers as well as transcriptomic evaluation of cellular responses to micropatterning. This work will enhance understanding of whether SMC abnormalities contribute to disease initiation and progression in PAH and will contribute to the broader effort of developing more complex models of pulmonary vascular disease.

This project aims to enhance EEG source localization by addressing electrode misplacement, which can possibly lead to errors in brain activity reconstruction. We developed a optimization algorithm on the quasi-static electromagnetic model to optimize electrode positions. Using the multipole expansion method, our model minimizes discrepancies between recorded and predicted EEG signals. Our work has applicability to many clinical scenarios, like stroke activity localization, and can enhance existing brain activity reconstruction protocols.

In this work, I propose an improved remeshing approach for particle method approximations. Particle approximation methods are a flexible tool for approximating solutions to nonlinear continuity equations, and are especially useful for aggregation-diffusion equations, which have important applications in fields ranging from modeling physical processes to neural networks. They work by decomposing functions into constituent parts, called particles. By tracking the motion and mass associated with each of these particles over time, we then use these to construct a high-resolution approximation to a desired solution. However, particle methods suffer from accuracy decay over time, necessitating remeshing (resetting particle positions) to maintain a useful approximation. It's important that the techniques used for this remeshing preserve existing structures, so that our approximation exhibits the same qualities as the true solution of the underlying equation. For instance, existing remeshing techniques often preserve conservation of mass, but not entropy decay. By combining remeshing techniques to periodically merge clustered particles and introduce new particles, I'm developing a method that maintains approximation accuracy and preserves structural properties. I present the results of the numerical analysis done using Python, as well as an implementation of the method using a finite-difference approach, which examines the approximation at various steps through time. This approach is expected to preserve the structure of the true solution within the particle method approximation, contributing to the development of robust particle methods for a broad class of partial differential equations.

Magnetoencephalography (MEG) is a powerful, noninvasive type of brain imaging that uses magnetic field readings from outside the skull to reconstruct the neuronal current sources that produce them in accordance with Maxwell’s equations. However, as these magnetic fields do not have unique current sources, algorithms are structured with constraints to guarantee the correct solution. In this project, we design a novel algorithm to reconstruct neural current sources. Using a cone-shaped beam with its vertex at the origin and a spherical-head model, we show we can reproduce any signal produced from within the cone using a current distribution on the cone’s surface, effectively allowing us to spatially localize the current source responsible for a given dataset of MEG measurements. I have employed this algorithm on an artificially produced dataset using MATLAB and assessed its effectiveness through reconstruction error analyses and visual techniques like heat maps. Future work will include testing the method on phantom-head data. We anticipate this algorithm is adaptable to non-spherical head geometries and cases involving multiple significant current sources, and we are working towards these advancements. Unlike other inverse methods, we expect our approach to assume minimal a priori knowledge about the brain’s conductivity profile, making it easier to implement in cases where detailed information about the subject's neural anatomy is limited.