Browsing by Author "Jose, Roberto Siasoco"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Publication Improving the Representation of Human Health Behavior in Spatial Agent-Based Models of Disease Spread(2023-08-03) Jose, Roberto Siasoco; Anderson, Taylor MGiven the significant threat of infectious respiratory diseases to global health, epidemiological models are essential tools for better understanding disease transmission. Of the many modeling approaches, Agent-Based Models (ABMs) are ideal for simulating disease spread because of their ability to examine and predict disease outcomes based on individual-level behaviors and interactions, which are key drivers of disease spread. However, many existing ABMs either ignore or generalize the behavioral component due to several challenges, such as the lack of data for informing agent behaviors, difficulties with implementing behavioral computational frameworks, or limited interdisciplinary collaboration between the broader ABM community and domain experts. The objective of the thesis is to advance the representation of human health behaviors in ABMs of infectious respiratory disease spread. To achieve this, a systematic literature review is conducted to assess the extent to which health behavior is modeled in existing ABMs of infectious respiratory disease spread. Building upon the findings from the literature review, a data-driven agent decision framework of health behaviors for spatial ABMs of disease spread is developed. The framework is then integrated into a geospatial ABM that simulates the spread of COVID-19 and mask-use behavior among the student population at George Mason University (GMU) during the Fall 2021 semester. The advancements made in this thesis will ultimately provide the public and decision-makers with greater insight into disease transmission, accurate predictions on disease outcomes, and preparation for future infectious respiratory disease outbreaks.Publication Study of Predictive Analysis of Hospital Mortality Using ECG Signals from Heart(2023-08-03) Jose, Roberto Siasoco; Wojtusiak, JanuszThe hearts electrical signals play an important role in its functioning to collect deoxygenated blood and pump oxygenated blood to the rest of the body. The abnormality in the hearts muscle can cause deficiencies in electrical signal generation or passage through heart valves which can lead to the imbalance of blood flow leading to a wide range of issues from a minor abnormality to a severe outcome such as death. Therefore, it is important to measure the strength of electrical signals on patients prone to heart diseases while admitted in a hospital or remotely using wearables. In this study, I experimented with different variations of ECG signals acquired on patients in a hospital setting to study the ability of advanced deep learning methodology vs. traditional signal processing to predict mortality of a patient in the hospital. The results are based on 198 patient cohort equally split between male and female. The results from deep learning are better than the traditional methods to predict patient mortality at hospital using ECG signals of heart.