Colorimetric assays are an important tool in point-of-care testing that offers several advantages to traditional testing methods such as rapid response times and inexpensive costs. A factor that currently limits the portability and accessibility of these assays are methods that can objectively determine the results of these assays. Current solutions consist of creating a test reader that standardizes the conditions the strip is under before being measured in some way. However, this increases the cost and decreases the portability of these assays. The focus of this study is to create a machine learning algorithm that can objectively determine results of colorimetric assays under varying conditions. To ensure the flexibility of a model to several types of colorimetric assays, three models were trained on the same convolutional neural network with different datasets. The images these models are trained on consist of positive and negative images of ETG, fentanyl, and HPV Antibodies test strips taken under different lighting and background conditions. A fourth model is trained on an image set composed of all three strip types. The results from these models show it is able to predict positive and negative results to a high level of accuracy.
Precise addition of agricultural inputs to maximize yields, especially in the face of environmental stresses, becomes important from the financial and sustainability perspectives. Given compounding factors such as climate change and disputed water claims in the American Southwest, the ability to build resistance against salinity stress becomes especially important. It was evaluated if an algal bio-fertilizer was able to remediate salinity stress in Solanum Lycopersicum. A hydroponic apparatus was employed, and data from Burge Environmental’s MiProbes™ both were able to demonstrate remediation. Future research could include determining the minimum dosage of algal fertilizer sufficient to induce this result, or the maximum concentration of salt that an algal treatment can provide a protective effect against.
Quantum computing is an emerging and promising alternative to classical computing due to its ability to perform rapidly complex computations in a parallel manner. In this thesis, we aim to design an audio classification algorithm using a hybrid quantum-classical neural network. The thesis concentrated on healthcare applications and focused specifically on COVID-19 cough sound classification. All machine learning algorithms developed or implemented in this study were trained using features from Log Mel Spectrograms of healthy and COVID-19 coughing audio. Results are first presented from a study in which an ensemble of a VGG13, CRNN, GCNN, and GCRNN are utilized to classify audio using classical computing. Then, improved results attained using an optimized VGG13 neural network are presented. Finally, our quantum-classical hybrid neural network is designed and assessed in terms of accuracy and number of quantum layers and qubits. Comparisons are made to classical recurrent and convolutional neural networks.
The debate around genetic engineering has permeated society for decades. A crucial aspect of this debate is the containment of genetically engineered organisms. This project outlines the three types of biocontainment and the conclusions drawn about each in the form of policy briefs. These briefs utilize case studies to sketch an overview of the current biocontainment paradigm in the United States. In addition, there is a brief discussing the major conclusions drawn from the case studies, as well as a brief containing useful definitions.