136879 research outputs found
Sort by
Scale-Up of Photocatalysis-Assisted Water Disinfection: Opportunities and Barriers
Get PDFDwindling supplies of readily available freshwater, coupled with the continuous growth of the human population, necessitates the purification and reuse of wastewater. This requires next-generation treatment processes, such as advanced oxidative processes (AOPs, e.g., photocatalysis), that offer the ability to remove both harmful microorganisms and other contaminants of emerging concern (CECs) from water. Building upon the laboratory-scale successes of AOPs for the large-scale remediation of water, however, requires process development. To accomplish this goal, a custom-built mobile platform capable of processing 15.14 liters (4 gallons) per minute of water is built and employed for studying the photocatalytic inactivation of Escherichia coli (E. coli) from water. This study indicated that the benchtop setup is as efficient as laboratory-scale studies in disinfecting water by photocatalysis. The study also indicated that catalyst recovery, regeneration and reuse is possible via a combination of gravity-assisted settling, centrifugation and air plasma treatment of the recovered photocatalysts
Assessment of Experiential and Artificial Intelligence Techniques in Agricultural Curriculum and Research Investigating Garlic Oil Dosage and Giraffe Fecal Microbiome
No full textThe agricultural industry is facing unprecedented changes resulting from the growing population which has prompted advances in technology. Farmers are tasked with the concern of improving productivity in an environmentally sustainably manner, which to date has involved the use of naturally derived feed additives (NFA). Advances in sequencing and educational technology have led to an increase in the opportunity to evaluate microbial populations and personalize student curriculum to include hands-on and artificial intelligence (AI) activities. In Study 1, using the in vitro gas production technique four different doses (0 mg/L; 20 mg/L; 320 mg/L and 620 mg/L) of garlic oil (GO) was tested to determine effects on methane production and rumen fermentation. Two 48-hr incubations were performed. Methane production was reduced (P ��� 0.01) by 27% with 320 mg/L GO. With higher concentrations of GO, the acetate:propionate ratio and fiber digestion were reduced (P ��� 0.01). The 320 mg/L dosage rate of GO seemed to favorably alter rumen fermentation parameters. In Study 2, fecal microbiome was characterized in giraffes voluntarily engaging in a guest feeding program (GFP). Giraffes (n = 6; Fort Worth Zoo) were divided into 3 groups dependent upon engagement frequency in GFP. Fecal samples were collected for 16S rRNA sequencing and behavior monitored. Giraffes that did not engage in GFP showed increased (P ��� 0.01) time spent foraging from hay-feeder sources, engaging in social interaction and object exploration. Giraffes that engaged in GFP showed greater concentrations (P ��� 0.01) of phyla Fibrobacterota, Firmicutes and Actinobacteriota. Therefore, giraffe engagement in GFP influenced behavior and composition of fecal bacteria and archaea communities. Undergraduate student perceptions (n = 310) of an experiential learning activity (ELA) in a livestock feed formulation course were investigated in Study 3. Participants completed a post-ELA demographic and 5-point Likert-scale survey. Most students agreed or strongly agreed they had a positive experience with the ELA (53% and 32%, respectively) and felt it positively contributed to course knowledge. Results indicated a favorable outcome for implementing ELA in animal science curriculum. Study 4 will investigate undergraduate student perceptions towards an AI tool in upper and lower-level agricultural courses
Investigation of Virtual Memory Aware Data Structures
Get PDFVirtual Memory Aware toolkit is a library of data structures with specific purposes, allowing for significant speedups to be found through the elimination of assumptions. This toolkit is designed to be flexible within its specific purposes, and allows for complete customizability of parameters for each project. While the library will be continuously updated as more structures are added, the primary focus is that of the bag and the stable index set. Bags are containers that do not guarantee sequence ordering, and as such are enabled to provide faster removals in cases where the order does not matter. Stable index sets are linear probed hash tables that perform resize operations by allocating levels of ever-increasing user-defined size to add new information to, allowing for quicker resize functions and stable access of members via a returned handle. This is supported by a fingerprinting algorithm, by which a selection of the hash is kept alongside each inserted object in order to allow for more efficient duplicate checking. While these algorithmic improvements allow for stronger use cases, they are not solely responsible for this. Following the theme of elimination of assumptions, instead of using malloc for memory allocation, VMA toolkit utilizes mmap, handling the memory allocation directly. By skipping over malloc, which contains logic to allow for generalized use cases, VMA toolkit is designed for high performance and large data set use cases, meaning that the memory allocation can be optimized for the specific use cases and create quicker allocation times throughout. Notably, handling allocation directly also means that the program can direct the kernel on how to handle the pages it has received, once again allowing for significant speedups compared to the standard implementation. Our experiments indicate that our VMA implementation of bag performs faster in all use cases than bag, demonstrating its utility for general purpose storage of data. Our data also shows that our stable hash table performs much faster than its current alternative, the naive index set, and performs comparably to, although not faster than, the standard hash table in all but specific delete cases, where the stable traits of the structure allows for gains over the standard hash table
The Colocalizationator: A Novel ImageJ Plugin for Semi-automated Cell Colocalization Analysis
Get PDFColocalization refers to the spatial overlap between two or more distinct fluorophores on the same cell or structure within a tissue sample and is frequently used to determine cellular identity. For example, in a section of spinal cord tissue, if a cell is positive for both ChAT and NeuN immunoreactivity, the cell is a motor neuron. For ex vivo brain or spinal cord tissue samples, colocalization between two or more image channels is typically assessed by manual (hand-counting) methods, which can be painstaking and time-consuming, or via semi-automatic methods with the help of software. This process can be optimized by using image analysis programs such as ImageJ to compare regions of interest (ROIs) to identify colocalized areas across multiple channels. However, despite a majority of the work being delegated to the computer, a significant portion of this process still requires manual input of functions by the researcher, which can end up taking hours of work for multiple images. Hence, there is a great need for software that can perform colocalization analysis with minimal input from the user that can generate high-fidelity data compared to manual quantification. Previously published literature on various plugins and software with vastly different modes of quantification have shown that automation, when programmed successfully, can lead to powerful tools that quickly yield extremely accurate and useful data for analysis. However, such literature has also suggested that highly specialized and specific quantification software often lead to errors and inaccuracies if the images' conditions are not ideal. To account for such variability and discrepancy in image types and environments, it is necessary to devise software that allows customization capable of catering to the optimum quantification settings. This led us to the creation and development of a fully automated plugin that allows the user to simply enter a few settings and step away while the computer does the work, regardless of the biological environment or complexity of the images
