Evaluation of and diagnosis based on core needle biopsies at present requires trained pathologists on site for both sample manipulation and analysis of tissue structures. Microfluidic lab-on-chip architectures have been studied for automating cell-level pathology for decades, and the CoreView system developed in Eric Seibel’s lab applies similar technologies at the millimeter scale to tissue level analysis. Pulsatile flow has proven to be a reliable means by which to transport tissue samples without damage or adhesion to the flow channels, and high resolution microscope images can be taken in glass-covered channels for analysis by a remote pathologist or image processing system. I am developing a low cost and manufacturable on chip device to accurately cut and sort these tissue samples for further processing. To avoid unnecessary complexity and unreliability, compliant elastomeric actuators will be employed to actuate an on-chip knife which also acts as a valve controlling transport flow routing. This integrated compact device will achieve all of these goals in a system that can be mass-produced for low-cost and disposability if required for sterility. Such capabilities will enable tumor-rich regions to be sampled for genomic analyses that allow precision therapy, making cancer a treatable disease.

The blood-brain barrier (BBB) is a highly specialized interface of brain microvascular endothelial cells (BECs). The main functions of BECs are to protect the brain against exposure to harmful substances in the blood, and to transport and secrete nutrients and other molecules that support normal brain functions. BECs can also alter their functions in response to signals from the brain or blood compartments. GLUT-1 is the predominant transporter at the BBB that regulates glucose entry into the brain, and does so by a mechanism of facilitated diffusion, which permits glucose transport in the blood-to brain or brain-to-blood direction. GLUT-1 dysfunction occurs in and may contribute to Alzheimer’s disease.The main hypothesis is that GLUT-1 malfunction in the BBB occurs as it deteriorates with age, which causes errors in other regulation methods of the BBB leading to the development of Alzheimer's. However, it is difficult to study mechanisms of GLUT-1 dysfunction at the BBB specifically because of the involvement of multiple cell types that regulate glucose uptake into the brain in vivo. In my research, I have utilized a model of iPSC-derived brain endothelial cells (iBECs) in order to study aspects of GLUT-1 regulation in an in vitro model of the BBB. Currently, my main findings have shown that the GLUT1 transporter is functional in our system, and that GLUT1 protein expression is increased and glycolytic enzymes are decreased when iBECs switch from a proliferative state to a quiescent state. Glucose transport in the blood-to-brain direction is decreased in the quiescent state.  Additionally, high glucose exposure causes downregulation of glucose transport.  My future studies aim to determine whether proliferating vs. quiescent iBECs respond differently to high or low concentrations of glucose. In summary, we have shown that iBECs are a translatable and effective model that allows us to further investigate the regulation of GLUT-1 at the BBB to further understand the physiological mechanisms involved in Alzheimer's Disease. 

The goal of this project is to develop a dynamically evolving connectionist model that more closely resembles the brain through its information-processing. Over the years, AI has shifted from the first generation of feedforward systems to the use of recurrent or convolutional Neural Networks. The third and newest generation of AI models, the brain-based models, and the Spiking Neural Network (SNN), attempts to bridge the gap between Neuroscience and ML using biologically realistic models like Θ-model, LIF, Izhikevich, HR, HH. These models, however, are still a black box leaving very little control or understanding on the learning process within the system without the access to the inner structure of the network. In addition, these systems are highly inefficient, slow, and very complex due to the limitations imposed by the hardware and explicit simulation of partial differential equations. Real world problems require “flexible learning and dynamically adaptive connectionist systems” that are capable to adapt and accommodate new input in real time. Current solutions have focused on varying the weights within a system rather than focusing on how connections within the system are formed. Based on our understanding from organismal brain structures, our approach, called biomimetic information codec, .bic, is a morphologically-adaptive coding hierarchical network that form in accordance with energy minimization - driven by dissipation of "heat" generated by the training data - constructing cortices and connectome for processing of information. My first objective herein is to quantitatively compare detailed structures between biological (fly brain) and .bic. networks using a random matrix approach.

Neurons in the brain are dynamical in nature, maintaining constantly changing states. Neurons modulate voltage based on input currents and produce spikes when the voltage exceeds a certain threshold. Additional dynamics after spiking, called evoked after-spike currents, are important for computation and memory over time scales. The diversity of neuronal dynamics and the variability in parameters underlying them give rise to rich and varied dynamics across networks. We hypothesize that the complexity and diversity of biological dynamics in the brain play a critical role in predictive coding of temporally complex systems, and that diverse forms of after-spike currents enable computation over variable timescales. Current artificial neural networks (ANNs), that emulate the structure of biological neural networks, successfully learn relationships between static patterns but have difficulty learning dynamic patterns that change over time. We aim to incorporate complex biological dynamics and diversity in ANNs and thereby systematically explore the function of such dynamics in network computation and learning. To this end, we construct ANNs that express biologically realistic dynamics, developing methods to learn dynamics-generating parameters, such as membrane capacitance and threshold, in individual neurons. Theoretically, diverse dynamics of individual neurons will enable even more complex dynamics when combined in networks and may improve performance on tasks requiring computation over complex timescales, such as determining actions based on temporal patterns of cues. Hence, we hypothesize that when trained on temporally challenging tasks, our networks will learn diverse dynamics across neurons. We present the diversity of parameters learned and the resulting distribution of firing patterns and compare performance between our neural networks and traditional networks that only learn connection weights. This research will inform learning methods for training novel biologically inspired neural networks and will also shed light on the physiological role of diversity in the brain.

Medical tape is used to hold essential devices to the skin for long periods of time. Unfortunately, without means for safe removal, these strong adhesives are painfully removed from the skin, often resulting in medical adhesive-related skin injuries (MARSI). Initial stakeholder interviews have indicated that medical tape removal is painful for the patient, and causes significant anxiety for nurses and caregivers. A 2015 study showed 98.6% of nurses considered skin tears common, occurring in 15% of senior patients and 17% of neonatal patients. A medical tape that offers high adhesion with means for safe removal is needed to eliminate MARSI and increase quality of care. UnTape addresses this need by providing a medical tape that has high adhesion during use but allows for easy and injury-free removal, by simply heating the tape for a few seconds with a heat pack prior to removal. The result is a rapid reduction of the force needed to remove the tape from the patient’s skin without risking MARSI. The tape is formulated with pressure-sensitive adhesive (PSA) that contains an embedded temperature-responsive additive (TRA). The additive will migrate to the surface of the tape upon heating, and melt in the range of 38-43°C, forming puddles that disrupt the adhesive and skin interface. The TRA is selected with a melting temperature that is high enough to avoid accidental peel strength reduction during fever, but below the pain threshold (45-47°C) for skin. Different additives have exhibited over a 50% reduction in peel force. This work focuses on optimization of product definition to yield consistent in vitro testing results, paving the way to clinical studies. The unique properties of UnTape allow for stronger skin adhesion for critical medical devices while eliminating the risk of MARSI upon removal, reducing nurse and patient stress, and providing higher quality medical care.