The posters presented here represent a small slice of this scholarship, but give some sense of the breadth and depth of the work that these students and faculty were able to carry on at a very high level, despite being scattered across the globe. We hope you'll enjoy reading and viewing the results of their hard work, and take advantage of the comments (while they're available) to ask and answer questions.
Kidney stone (KS) disease is the most common disease of the urinary tract. Over 1.3 million individuals visit emergency departments (ED) for stone-related symptoms, which is projected to increase. Since most lab tests are misinterpreted, leading to an incorrect diagnosis, diagnostic predictive tools are useful to most accurately calculate an individual's likelihood, or risk, of having KS disease. We investigate which predictive model would yield the highest accuracy and sensitivity by extracting information from a clinical database containing patient data and utilize various machine learning and statistical analysis to analyze the performance of each model. With using confusion matrices, logistic regression, and ROC curves, it has been found that including all domains in a predictive model (demographics, laboratory tests, ICD-9 diagnoses, etc.) demonstrated the highest performance.
The Undergraduate ALFALFA Team (UAT) Groups Project is investigating the influence of the environment on galaxies in group environments. Galaxies in clusters show evidence for a variety of environmental effects. The UAT team has chosen groups from the RASSCALS sample (groups defined using the Rosat X-ray satellite, Mahdavi et al. 2000). Galaxies within 2 degrees of group center and 3 sigma of the central velocity of the group were targeted for Hα observations. We cross-matched the 52 RASSCALS groups and clusters the UAT has observed to the Tempel et al. (2017) group catalog based on Sloan Digital Sky Survey observations, finding 26 that matched within 10 arcminutes of the RASSCALS group. Here we compare properties of these groups.
Feasibility of an in-home remote exercise intervention for older adults with mild cognitive impairment during the COVID-19 pandemic: a pilot study
With the prevalence of dementia and neurocognitive decline on the rise, older adult populations have begun seeking out non-pharmacological methods to prevent or ameliorate their symptoms. Specifically, prior research has shown greater cognitive benefits from simultaneous exercise and cognitive stimulation than exercise alone, particularly for patients with mild cognitive impairment (MCI). In response to the COVID-19 pandemic, we conducted a pilot study to test the feasibility of a remote, home-based, exercise intervention for older adults. For three months, six older adults engaged in integrative physical and cognitive exercise via a neuro-exergame called the interactive Physical and Cognitive Exercise System (iPACES v2.0), which involves pedaling an under-desk elliptical while playing an interactive video game. Participants completed the Montreal Cognitive Assessment (MoCA) before and after the intervention, and exit interviews were conducted via Zoom at the conclusion of the study. While participants found the neuro-exergame to be entertaining, several setup and technical difficulties arose that made their overall experience more difficult. Our findings suggest that a more effective intervention would involve more user-friendly features that allow for greater ease of use by older adults with MCI.
As the number of applications to universities across America has increased in recent years, the ability to predict which applicants intend to enroll has become integral to the success of undergraduate institutions. Using data analysis to predict matriculation enables admissions departments to make more accurate forecasts of enrollment and finances. We use machine learning techniques to estimate the likelihood of enrollment for each applicant in a given year. These probabilities of enrollment are, in turn, used to calculate the aggregate incoming class profile, allowing an admissions department to tailor their admission decisions accordingly. The study experiments with various machine learning algorithms: logistic regression, random forest and XGBoost models are all built and tested. We find that the XGBoost algorithm consistently outperforms other algorithms in predicting enrollment. We use a random subset of 2013-18 data for training, and the remainder of the subset for validation. Academic strengths, financial offers, and applicant engagement all possess predictive power on enrollment in the model. We conclude the study by applying the model to the accepted students from the 2019 applicant class. This approach serves as a proxy for predicting the profile of an incoming class profile: the model uses the application data to create predictions of enrollment without knowledge of which applicants enrolled. We then compare the predicted class profile to the actual class profile to assess the model’s predictive accuracy.
Applications of Metal Nanoparticles and Comparison of Nanosheets with Gold and Nanosheets without Gold using Atomic Force Microscopy
Nanoparticles have become common in many real world devices, including catalysts, data storage devices, sensors and solar cells. This increase in usage is due to the ability of the nanoparticles to easily assemble into two-dimensional structures with various useful optical and electronic properties. The goal of this summer project was to 1) research current applications of metal nanoparticles relevant to the gold nanoparticle-embedded nanosheets formed in our lab and 2) to use computer software to analyze atomic force microscopy (AFM) images of gold nanoparticle-embedded nanosheets formed in our lab this summer. In the lab, we are aiming to create and characterize stable planar gold nanoparticle materials, such as emulsions and nanosheets, by assembling the gold nanoparticles with peptoids at fluid surfaces. The characterization of these materials is accomplished through AFM, which is used to analyze the heights of the nanosheets and to determine whether or not they can be successfully formed both with and without gold nanoparticles. The AFM images obtained this summer were processed and analyzed using IgorPro so that the heights of the different materials could be obtained and compared. This project is based on work from the previous 2019-2020 academic year.
Remote administered in-home interactive Physical and Cognitive Exercise Study (iPACES v2.8) for youth on the Autism Spectrum: A feasibility pilot
The portable Interactive Physical and Cognitive Exercise System (iPACES, v 2.8), tablet, smart watch and an under-desk stationary elliptical will be sent to the home of a participant and used in-home for three months in this experiment. We aim to examine the effects of cognitive and physical stimulation, in adolescents on the autism spectrum. This three-week intervention of exergaming will add to the finding of previous studies that used a 20 minute bout of exergaming. Participants will receive the iPACES set-up and necessary testing materials via mail to remain consistent with the social distancing guidelines due to COVID-19. To begin the experiment, participants will be given a series of questionnaires and cognitive tasks, such as Stroop, Color Trails, and Digit Span via a video call on a remote video conference platform (e.g. Zoom) remaining consistent with the tele-health guidelines. Participants will then be asked to use the game and peddler combination 3-5 times weekly for 30-45 minutes. While this proposed study does not attempt to maximize participation in exercise, it aims to support the belief that exergaming produces a positive effect on cognitive functioning and the reduction of autism-related behaviors of repetition and physical self-stimulation. It will also indirectly investigate the possibility of the use of exergaming as a viable treatment for autism and the feasibility of a remote protocol.
The high-resolution imaging of the radar/ladar system is typically obtained by transmitting the wideband waveforms. These wide bandwidth waveforms are generated by modulating signals on to a frequency carrier which requires additional hardware. To counter this, we propose a semiconductor laser system that can generate a chaotic waveform with a frequency span of a few Giga-Hertz. The generated waveforms are pseudo-random and hence can yield sharp correlation peaks. These features are essential to identify multiple hotspots in close proximity.
A cataclysmic variable star system consists of two stars that are close enough that material can be transferred from one to the other through an accretion stream onto a circumstellar accretion disk. We observe variation in the light due to rotation of the accretion hotspots, variations of the material flow rate, and the change of view angle due to orbital motion of the binary. As part of the Center for Backyard Astrophysics (CBA), an amateur-professional collaboration, we chose high priority targets and observed them using the 0.6m telescope of El Sauce Observatory in Chile. We observed HP Lib on three nights in June and July and noted an overall brightening. We obtained photometry on NY Lup for four nights during a single week in July and detected a local minimum. Our results will be combined with those of other CBA observers and will be analyzed for variation on multiple timescales allowing a detailed interpretation in terms of physical changes of the systems.
The goal of this research is to develop a gesture control system prototype for a robot by utilizing Oculus Quest real-time hand tracking feature. The robot is currently run in a virtual simulated world. Users can control its movement by interacting with the control terminal that implemented in Unity VR. The objects in the virtual simulated world are recreated in Unity game world by a real-time generated mesh using the camera data from the robot.
Biomarker and cognitive improvements for MCI patients after neuro-exergaming: Pedal and play for brain health (iPACES v2.5 and 2.75)
There is a dementia epidemic that is affecting the older adult population, and researchers are exploring accessible ways to ameliorate the cognitive decline associated with Alzheimer’s Disease and related dementias. Non-pharmacological interventions such as interactive physical and cognitive exercise are being investigated to understand the physiological and cognitive effects in older adults. Twenty-seven older adults were enrolled in a single-bout neuro-exergaming intervention of the interactive Physical and Cognitive Exercise System (iPACESv2.0), a neuro-exergame that consists of pedaling an under-desk elliptical while playing an interactive video game. The study explores the cognitive and biomarker outcomes in participants with mild cognitive impairment (MCI). The intervention featured a neuropsychological battery and salivary analysis to measure changes in executive function and biomarkers associated with neuroplasticity. Analyses revealed a significant increase in executive function and salivary alpha-amylase in the MCI population, suggesting cognitive improvements occurred after the intervention. This study provides encouraging preliminary support for the use of interactive exergaming interventions as clinical treatments to ameliorate the cognitive decline associated with Alzheimer’s disease.
A model was developed to predict the amount of solar radiation incident on a solar collector’s surface. A Python based model tracked the sun’s location and determined how much energy would strike a surface at any orientation. As a demonstration several fixed tilt angles and a tracking case were simulated for Schenectady, NY. Results followed the expected trends. Future goals for the model are to confirm its results numerically with experimental data and use it to determine optimal arrangements for bifacial solar collectors.
The overarching goal of this research was to simultaneously design, manufacture, and test different mechanisms for robotic tick collection, along with designing and developing an electronics package and code that would allow the robot to behave in a semi-autonomous fashion. The tick collection mechanisms were designed with the intention of providing an optimal efficiency for physically collecting ticks, minimizing their impact on the overall maneuverability of the robot, being as cost effective as possible, and being as easy to replicate as possible. A total of seven different prototypes were developed and tested under real world conditions. Of these seven, four demonstrated very promising results regarding the functional requirements of the project. The electronics package and accompanying code are being developed with the intention of allowing the robot to autonomously navigate environments which would make manual control difficult, such as dense shrubbery that would impair a direct line of sight.
As of mid-2020, two major crises currently afflict the United States, overlapping and compounding one another: COVID-19 and racial injustice. Globally, as of August 4, 2020, there have been 18,142,718 confirmed cases of COVID-19, including 691,013 deaths, reported to WHO. Of that, the United States has had 4,629,459 confirmed cases with 154,226 deaths. This means that while the US comprises only 4.25% of the total world population, it makes up 25.5% of all cases and 22.3% of all deaths. The coronavirus is classified as a pandemic with a significant number of undetected, asymptomatic cases, as many people travel, interact and transmit the virus to others, leading to massive outbreaks. There is increased risk with increased age and underlying health conditions, but one pattern that has become clear in the US has been the disproportionate increased risk of contraction and death for BIPOC (Black, Indigenous, People of Color). Recently, The New York Times sued the CDC in order for them to reveal information that confirms drastic disparities in the impact of COVID-19 on African American, Latino and Native American communities. Latino and African-American residents of the United States have been three times as likely to become infected as their white neighbors, according to this new data. Why is this? Biological skin color does not affect one’s risk, but the systems we have in place, emphasizing racial inequity, definitely do. Due to long-standing systemic health and social inequities, racial minorities are at increased risk of getting sick and dying from COVID-19, according to the CDC. This research project investigates the COVID-19 health crisis in the US, how it is connected to racial injustice with health and social inequities placing racial minorities in disproportionate harm, on top of how the Trump administration’s actions/inactions have heightened these issues in such a way that the compounded crisis exposes the most severe, long-lasting and deadly consequences of the politics of structural racism.
Thermal Modelling of Aerogel-Based Washcoats for Three-Way Catalytic Conversion: Effects of Washcoat Composition and Thickness on Light-Off Time
Three-way catalytic converters typically consist of an encased cordierite honeycomb structure, coated with an alumina washcoat impregnated with platinum group metals (PGMS). While effective under normal operating conditions, there is typically a significant warm-up time during which the TWC is ineffective. Aerogels are nanoporous materials that have a large surface area, low density, and low thermal conductivity. The use of aerogel in place of the more dense and thermally conductive alumina washcoat might reduce the time needed to heat up the TWC and decrease overall pollutant emissions. This idea was investigated using a one-dimensional model to simulate heat transfer in a cordierite wall with three different coatings: silica aerogel washcoat, a catalytically active copper-alumina (CuAl) aerogel washcoat and a traditional alumina washcoat. Simulations were performed using a transient finite difference model in MATLAB and confirmed in Abaqus. The exhaust gas was assumed to flow over the surface at a temperature of 600 K with a heat transfer coefficient of 35 W/m2K, which is typical for a catalytic converter. Two different honeycomb structures were analyzed, 75-µm-thick and 150-µm-thick cordierite walls, to simulate 400 and 300 cells per square inch (CPSI) honeycombs. A nominal washcoat thickness of 20 µm was modelled, and the surface temperature of the washcoat in direct contact with the exhaust was analyzed over time. A number of scenarios were examined including: (a) the effect of washcoat properties; (b) the effect of the percent (0-100%) of the washcoat thickness in the total wall composition; (c) the effect of washcoat thickness (10-100 µm) for fixed cordierite thickness; and (d) the effect of a transient exhaust temperature. The results show that there is a large initial increase in surface temperature for the aerogel-based washcoats (compared to that of the alumina) which then levels off as the heat penetrates into the cordierite layer. The aerogel-based washcoat reaches light-off temperature (the temperature at which the TWC converts 50% of the pollutants) 2-3 sec faster than the alumina washcoat for the 20-µm layer. For a 100-µm layer, the aerogel washcoats reach light off 13-14 sec faster. Increasing the percentage of aerogel in the total wall composition significantly reduces the time to light-off by 29 sec (300 CPSI honeycomb) and by 16 sec (400 CPSI); however, light-off time increases with washcoat layer thickness. Therefore, using a thinner coating is better under all conditions modelled. When using a more realistic model with an exhaust temperature that increases with time, similar trends are observed. Although there are a number of challenges associated with using an aerogel-based washcoat, these results indicate that their use could reduce TWC light-off time. Finding a way to maintain the initial surface temperature rise would allow even shorter light-off times to be achieved.
Dementia is a symptom of many neurodegenerative disorders that have a heavy economic burden. In this research, we are building a diagnostic model utilizing a big data approach to analyze clinical records of demographics, diagnoses, vital signs, and other clinical information. We used a large database from one hospital in which thousands of patients were admitted over a span of a decade. We extracted multiple variables which would be relevant for diagnosing patients with dementia and analyzed them through machine learning models.
An investigation of physical and mental disability in theatre; the importance of accessibility and inclusion, and the methods by which to get there.
By Prof. Anderson’s guidance, I’ve implemented an algorithm that can reduce the computational complexity of Matrix Multiplication.