Permanent URI for this collection
Browse
Recent Submissions
Publication Optimal Mammography Schedule Estimates Under Varying Disease Burden, Infrastructure Availability, and Other Cause Mortality: A Comparative Analyses of Six Low- and Middle- Income Countries(2020-09) Shifali, ShifaliLow-and-middle-income countries (LMICs) have a higher mortality-to-incidence ratio for breast cancer compared to high-income countries (HICs) because of late-stage diagnosis. Mammography screening is recommended for early diagnosis, however, current screening guidelines are only generalized by economic disparities, and are based on extrapolation of data from randomized controlled trials in HICs, which have different disease burdens and all-cause mortality compared to LMICs. Moreover, the infrastructure capacity in LMICs is far below that needed for adopting current screening guidelines. This study analyzes the impact of disease burden, infrastructure availability, and other cause mortality on optimal mammography screening schedules for LMICs. Further, these key features are analyzed under the context of overdiagnosis, epidemiologic/clinical uncertainty in pathways of the initial stage of cancer, and variability in technological availability for diagnosis and treatment. It uses a Markov decision process (MDP) model to estimate optimal schedules under varying assumptions of resource availability, applying it to six LMICs. Results suggest that screening schedules should change with disease burden and life-expectancy. For countries with similar life-expectancy but different disease burden, the model suggests to screen age groups with higher incidence rates. For countries with similar incidence rate and different life expectancy, the model suggests to screen younger age groups for countries with lower life-expectancy. Overdiagnosis and differences in screening technology had minimal impact on optimal schedules. Optimality of screening schedules were sensitive to epidemiologic/clinical uncertainty. Results from this study suggest that, instead of generalized screening schedules, those tailored to disease burden and infrastructure capacity could help optimize resources. Results from this study can help inform current screening guidelines and future health investment plans.Publication MARKOV DECISION PROCESS APPROACH TO STRATEGIZE NATIONAL BREAST CANCER SCREENING POLICY IN DATA-LIMITED SETTINGS(2019-09) Deshpande, VijetaEarly diagnosis is a promising strategy to reduce premature mortalities and for optimal use of resources. But the absence of mathematical models specific to the data settings in LMIC’s impedes the construction of economic analysis necessary for decision-makers in the development of cancer control programs. This thesis presents a new methodology for parameterizing the natural history model of breast cancer based on data availabilities in low and middle income countries, and formulation of a control optimization problem to find the optimal screening schedule for mammography screening, solved using dynamic programming. As harms and benefits are known to increase with the increase in the number of lifetime screens, the trade-off was modeled by formulating the immediate reward as a function of false positives and life-years saved. The method presented in thesis will provide optimal screening schedules for multiple scenarios of Willingness to Pay (numeric value assigned for each life-year lived), including the resulting total number of lifetime screens per person, which can help decision-makers evaluate current resource availabilities or plan future resource needs for implementation.Publication ADVANCED VIRTUAL REALITY HEADSET BASED TRAINING TO IMPROVE YOUNG DRIVERS’ LATENT HAZARD ANTICIPATION ABILITY(2019-09) Agrawal, RaviDriving safety among young novice driver is one of the largest concern in the transportation domain. Many Paper-based or PC- based training program have been developed over the years to train the young novice driver to improve their driving skills (Hazard Anticipation). This training programs does help young novice driver to improve their situational awareness and so the hazard anticipation skills. But, there is one common problem with most of the currently available training programs. They are not very immersive, because such training program mostly provide plain view of the training scenario’s along with some description about the scenario and the subject trained in such training method needs to translate the provided knowledge in the plain view into the real-world driving. An Advanced training program on risk awareness and perception was developed and evaluated in Oculus rift platform. The primary objective is to train the young novice driver in the Virtual reality headset based risk awareness and perception training program and evaluate the trained driver in the driving simulator against the placebo trained young novice driver. The Virtual reality headset based risk awareness and perception training program (V-RAPT) is based on 3M Error-based Training approach where the driver will have 80 horizontal degrees’ and 90 vertical degrees’ field of view. Thirty-six drivers will receive training in the respective training methods- V-RAPT (Virtual reality headset based risk awareness and perception training), RAPT (PC- based risk awareness and perception training) and placebo training. Twelve young novice driver trained in the V-RAPT group will served as experimental group. Twenty-four other young novice will receive training in the RAPT and Placebo training respective will serve as control group. After training all three-group trained driver will be evaluated in the advanced driving simulator and the eye movement of the all thirty-six participants are recorded and measured. Vehicle measures such as acceleration, velocity and brake position is also recorded. The drivers’ score will based on whether or not their eye-fixations indicated recognition of potential risks in different high risk driving situations. The evaluation driver included six scenarios used in the V-RAPT training (near transfer scenarios) and four scenarios that were not used in the V-RAPT training (far transfer scenarios). Drivers who received the V-RAPT training are expected to drive more safely than the drivers who received either training. The V-RAPT trained drivers are expected to glance on regions (Hazard anticipation) where potential risks might appear than the drivers’ trained in the RAPT and Placebo training method. Further, The V-RAPT trained drivers are expected have slower average velocity and better brake position (Hazard mitigation) are compared to the driver trained in the other two training method.Publication A Computational Simulation Model for Predicting Infectious Disease Spread using the Evolving Contact Network Algorithm(2019-05) Munkhbat, BuyannemekhCommonly used simulation models for predicting outbreaks of re-emerging infectious diseases (EIDs) take an individual-level or a population-level approach to modeling contact dynamics. These approaches are a trade-off between the ability to incorporate individual-level dynamics and computational efficiency. Agent-based network models (ABNM) use an individual-level approach by simulating the entire population and its contact structure, which increases the ability of adding detailed individual-level characteristics. However, as this method is computationally expensive, ABNMs use scaled-down versions of the full population, which are unsuitable for low prevalence diseases as the number of infected cases would become negligible during scaling-down. Compartmental models use differential equations to simulate population-level features, which is computationally inexpensive and can model full-scale populations. However, as the compartmental model framework assumes random mixing between people, it is not suitable for diseases where the underlying contact structures are a significant feature of disease epidemiology. Therefore, current methods are unsuitable for simulating diseases that have low prevalence and where the contact structures are significant. The conceptual framework for a new simulation method, Evolving Contact Network Algorithm (ECNA), was recently proposed to address the above gap. The ECNA combines the attributes of ABNM and compartmental modeling. It generates a contact network of only infected persons and their immediate contacts, and evolves the network as new persons become infected. The conceptual framework of the ECNA is promising for application to diseases with low prevalence and where contact structures are significant. This thesis develops and tests different algorithms to advance the computational capabilities of the ECNA and its flexibility to model different network settings. These features are key components that determine the feasibility of ECNA for application to disease prediction. Results indicate that the ECNA is nearly 20 times faster than ABNM when simulating a population of size 150,000 and flexible for modeling networks with two contact layers and communities. Considering uncertainties in epidemiological features and origin of future EIDs, there is a significant need for a computationally efficient method that is suitable for analyses of a range of potential EIDs at a global scale. This work holds promise towards the development of such a model.Publication Effects of Task Load on Situational Awareness During Rear-End Crash Scenarios - A Simulator Study(2019-05) Nair, RajivThe current driving simulator study investigates the effect of 2 distinct levels of distraction on a drivers’ situational awareness and latent and inherent hazard anticipation. In this study, rear-end crashes were used as the primary crash configuration to target a specific category of crashes due to distraction. The two types of task load used in the experiment was a cognitive distraction (mock cell-phone task) & visual distraction (I-pad task). Forty-eight young participants aged 18-25 years navigated 8 scenarios each in a mixed subject design with task load (cognitive or visual distraction) as a between-subject variable and the presence/absence of distraction representing the within-subject variable. All participants drove 4 scenarios with a distraction and 4 scenarios without any distraction. Physiological variables in the form of Heart rate and heart rate variability was collected for each participant during the practice drives and after each of the 8 experimental drives. After the completion of each experimental drive, participants were asked to fill up a NASA TLX questionnaire which quantifies the overall task load experienced by giving it a score between 1 and 100, where higher scores translate to higher perceived task load. Eye-movements were also recorded for the proportion of latent and inherent hazards anticipated and mitigated for all participants. Standard vehicle data (velocity, acceleration & lane offset) were also collected from the simulator for each participants’ each drive. Analysis of data showed that there was a significant difference in velocity, lane offset and task load index scores across the 2 groups (between-subject factors). The vehicle data, heart rate data and TLX data was analyzed using Mixed subject ANOVA. There was also a logistic regression model devised which showed significant effects of velocity, lane offset, TLX scores and age on a participants’ hazard anticipation abilities. The findings have a major practical implication in reducing drivers’ risk of fatal, serious or near crashes.Publication THE EVACUATION PROBLEM IN MULTI-STORY BUILDINGS(2019-02) Cung, Quang HongThe pressure from high population density leads to the creation of high-rise structures within urban areas. Consequently, the design of facilities which confront the challenges of emergency evacuation from high-rise buildings become a complex concern. This paper proposes an embedded program which combines a deterministic (GMAFLAD) and stochastic model (M/G/C/C State Dependent Queueing model) into one program, GMAF_MGCC, to solve an evacuation problem. An evacuation problem belongs to Quadratic Assignment Problem (QAP) class which will be formulated as a Quadratic Set Packing model (QSP) including the random flow out of the building and the random pairwise traffic flow among activities. The procedure starts with solving the QSP model to find all potential optimal layouts for the problem. Then, the stochastic model calculates an evacuation time of each solution which is the primary decision variable to figure the best design for the building. Here we also discuss relevant topics to the new program including the computational accuracy and the correlation between a successful rate of solving and problems’ scale. This thesis examines the relationship of independent variables including arrival rate, population and a number of stories with the dependent variable, evacuation time. Finally, the study also analyzes the probability distribution of an evacuation time for a wide range of problem scale.Publication Evaluation and Validation of Distraction Detection Algorithms on Multiple Data Sources(2018-09) Mehrotra, ShashankThis study aims to evaluate algorithms designed to detect distracted driving. This includes the comparison of how efficiently they detect the state of distraction and likelihood of a crash. Four algorithms that utilize measures of cumulative glance, past glance behavior, and glance eccentricity were used to understand the distracted state of the driver and were validated on two separate data sources (i.e., simulator and naturalistic data). Additionally, an independent method for distraction detection was designed using data mining methods. This approach utilized measures like steering degree, lane offset, lateral and longitudinal velocity, and acceleration. The results showed a higher likelihood of distracted events when cumulative glances were considered. However, the state of distraction was observed to be higher when glance eccentricity was added. Additionally, it was observed that glance behavior using the four legacy algorithms were better detectors of the state of distraction as compared to the data mining method that used vehicular measures. This research has implications in understanding the state of distraction, predicting the power of different methods, and comparing approaches in different contexts (naturalistic vs simulator). These findings provide the fundamental building blocks towards designing advanced mitigation systems that give drivers feedback in instances of high crash likelihood.Publication Impact of Perceptual Speed Calming Curve Countermeasures On Drivers’ Anticipation & Mitigation Ability – A Driving Simulator Study(2018-09) Valluru, KrishnaA potential factor for curve accidents are anticipatory skills. Horizontal curves have been recognized as a significant safety issue for many years. This study investigates the impact and effectiveness of three curve based perceptual speed calming countermeasures (advanced curve warning signs, chevron sign, and heads-up display(HUD) sign) on drivers’ hazard anticipation and mitigation behavior across both left and right-winding curves, and sharp (radius 200m) and flat (radius 500m) curves. Experimental results show that the speed and lateral control in the horizontal curves differed with respect to curve radii, direction, and the type of countermeasure presented. These differences in behavior are probably due to curve-related disparities, the type of perceptual countermeasure, and the presence of hazard at the apex of the curve. HUD is found to be effective at not only reducing the drivers’ speed in the curve, but also improve the latent hazard anticipation ability of the driver at the apex of the curve. Flat and sharp curves with indications of a safety problem were virtually developed in the simulator as representative as possible without upsetting the simulator’s fidelity. 48 participants were recruited for this study between the age range of 18 and 34, and driving experience range was from 0.25 to 17.75 years.Publication The Application of Usability Engineering Methods to Evaluate and Improve a Clinical Decision Support System(2018-05) DeSotto, KristineDelays in the process of diagnosing and treating cancer are common and lead to confusion and undesirable outcomes. Care coordinators are often embedded within the system of care to manage follow-up care. Electronic and real-time reminder systems can be used to support the care coordinator’s work, but electronic health record (EHR) usability is known to be poor. This study, completed in collaboration with the Department of Veterans Affairs (VA) Connecticut Healthcare System, evaluated the Cancer Coordination and Tracking System (CCTS), an EHR-linked, web-based tool for cancer care management. A set of expert-driven and user-driven usability engineering methods was applied to comprehensively identify and analyze usability problems within the system. Ten current CCTS users were engaged in the study to help identify problem. 101 (62.3%) problems were identified through expert-driven methods, 56 (34.6%) were identified by user-driven methods, and 5 (3.1%) were identified through both types of methods. The list of 162 unique problems were prioritized and twelve high priority problems were highlighted. Design recommendations were developed to address each of these high priority problems.Publication Does the Elicitation Mode Matter? Comparing Different Methods for Eliciting Expert Judgement(2018-05) Cruickshank, ClaireAn expert elicitation is a method of eliciting subjective probability distributions over key parameters from experts. Traditionally an expert elicitation has taken the form of a face-to-face interview; however, interest in using online methods has been growing. This thesis compares two elicitation modes and examines the effectiveness of an interactive online survey compared to a face-to-face interview. Differences in central values, overconfidence, accuracy and satisficing were considered. The results of our analysis indicated that, in instances where the online and face-to-face elicitations were directly comparable, the differences between the modes was not significant. Consequently, a carefully designed online elicitation may be used successfully to obtain accurate forecasts.Publication Explorations into Machine Learning Techniques for Precipitation Nowcasting(2017-02) Nagarajan, AdityaRecent advances in cloud-based big-data technologies now makes data driven solutions feasible for increasing numbers of scientific computing applications. One such data driven solution approach is machine learning where patterns in large data sets are brought to the surface by finding complex mathematical relationships within the data. Nowcasting or short-term prediction of rainfall in a given region is an important problem in meteorology. In this thesis we explore the nowcasting problem through a data driven approach by formulating it as a machine learning problem. State-of-the-art nowcasting systems today are based on numerical models which describe the physical processes leading to precipitation or on weather radar extrapolation techniques that predict future radar precipitation maps by advecting from a sequence of past maps. These techniques, while they can perform well over very short prediction horizons (minutes) or very long horizons (hours to days), tend not to perform well over medium horizons (1-2 hours) due to lack of input data at the necessary spatial and temporal scales for the numerical prediction methods or due to the inability of radar extrapolation methods to predict storm growth and decay. Given that water must first concentrate in the atmosphere as water vapor before it can fall to the ground as rain, one goal of this thesis is to understand if water vapor information can improve radar extrapolation techniques by giving the information needed to infer growth and decay. To do so, we use the GPS-Meteorology technique to measure the water vapor in the atmosphere and weather radar reflectivity to measure rainfall. By training a machine learning nowcasting algorithm using both variables and comparing its performance against a nowcasting algorithm trained on reflectivity alone, we draw conclusions as to the predictive power of adding water vapor information. Another goal of this thesis is to compare different machine learning techniques, viz., the random forest ensemble learning technique, which has shown success on a number of other weather prediction problems, and the current state-of-the-art machine learning technique for images and image sequences, convolutional neural network (CNN). We compare these in terms of problem representation, training complexity, and nowcasting performance. A final goal is to compare the nowcasting performance of our machine learning techniques against published results for current state-of-the-art model based nowcasting techniques.Publication Mining High Impact Combinations of Conditions from the Medical Expenditure Panel Survey(2023-09) Mohan, ArjunThe condition of multimorbidity — the presence of two or more medical conditions in an individual — is a growing phenomenon worldwide. In the United States, multimorbid patients represent more than a third of the population and the trend is steadily increasing in an already aging population. There is thus a pressing need to understand the patterns in which multimorbidity occurs, and to better understand the nature of the care that is required to be provided to such patients. In this thesis, we use data from the Medical Expenditure Panel Survey (MEPS) from the years 2011 to 2015 to identify combinations of multiple chronic conditions (MCCs). We first quantify the significant heterogeneity observed in these combinations and how often they are observed across the five years. Next, using two criteria associated with each combination -- (a) the annual prevalence and (b) the annual median expenditure -- along with the concept of non-dominated Pareto fronts, we determine the degree of impact each combination has on the healthcare system. Our analysis reveals that combinations of four or more conditions are often mixtures of diseases that belong to different clinically meaningful groupings such as the metabolic disorders (diabetes, hypertension, hyperlipidemia); musculoskeletal conditions (osteoarthritis, spondylosis, back problems etc.); respiratory disorders (asthma, COPD etc.); heart conditions (atherosclerosis, myocardial infarction); and mental health conditions (anxiety disorders, depression etc.). Next, we use unsupervised learning techniques such as association rule mining and hierarchical clustering to visually explore the strength of the relationships/associations between different conditions and condition groupings. This interactive framework allows epidemiologists and clinicians (in particular primary care physicians) to have a systematic approach to understand the relationships between conditions and build a strategy with regards to screening, diagnosis and treatment over a longer term, especially for individuals at risk for more complications. The findings from this study aim to create a foundation for future work where a more holistic view of multimorbidity is possible.Publication Systematic Review of Driver Distraction in the Context of Advanced Driver Assistance Systems (ADAS) & Automated Driving Systems (ADS)(2022-09) Hungund, Apoorva PramodAdvanced Vehicle Systems promise improved safety and comfort for drivers. Steady advancements in technology are resulting in increasing levels of vehicle automation capabilities, furthering safety benefits. In fact, some of these vehicle automation systems are already deployed and available, but with promised benefits, such systems can potentially change driving behaviors. There is evidence that drivers have increased secondary task engagements while driving with automated vehicle systems, but there is a need for a clearer scientific understanding of any potential correlations between the use of automated vehicle systems and potentially negative driver behaviors. Therefore, this thesis aims to understand the state of knowledge on automated vehicle systems and their possible impact on drivers’ distraction behaviors. I have conducted two systematic literature reviews to examine this question. This thesis reports these reviews and examines the effects of secondary task engagement on driving behaviors such as take-over times, visual attention, trust, and workload, and discusses the implications on driver safety.Publication Robustness of Supply Chain Synchronization Strategies(2021-09) Frere, AndrewModern manufacturing systems dealing with complex assemblies with large numbers of parts present particular challenges in the realm of supply chain management. Complex assemblies, such as those found in aerospace and automobile manufacturing, require thousands of parts to come together at the right time for final assembly. The large number of parts, often coming from hundreds of suppliers, combined with unreliable delivery times and high cost of many of these components can lead to incredibly high inventory costs and assembly delays. Typically, variable delays in part delivery are compensated for by either keeping a buffer of safety stock or a time buffer on the planned lead time of a component. In this thesis, we study the performance of various buffering strategies across a large range of practical scenarios in an effort to identify dominant, robust strategies and how their performance is impacted by the various parameters that define the system. The major conclusion is that aggressive part buffering consistently results in not only better delivery performance but also significant inventory reduction across all settings for assemblies with more than 500 parts.Publication Comparing and Improving the Design of Physical Activity Data Visualizations(2021-09) Frackleton, Peter MHeart disease is a leading cause of death in the United States, and older adults are at highest risk of being diagnosed with heart disease. Consistent physical exercise is an effective means of deterring onset of heart disease, and physical activity tracking devices can inspire greater activity in older adults. However, physical activity tracking device abandonment is quite common due to limitations on what can be learned from the activity data that is collected. Better data visualization of physical data presents an opportunity to surpass these limitations. In this thesis, a task-based human subject study was performed with three different data visualizations to gain insight into how the format of physical activity data visualizations impact older adults’ abilities to infer meaning from physical activity data. Participants (n = 30) interacted with a prototype data visualization as well as two data visualizations from popular fitness tracking applications (Fitbit and Strava) and used these visualizations to complete 11 tasks. Results from these tasks show each visualization was able to facilitate users answer some task questions effectively, though no visualizations exhibited strong performance across all tasks. From the successes and shortcomings of each visualization, three key design recommendations for the design of data visualizations for physical activity data were made: 1) make exact values available, 2) summarize data at multiple timescales, and 3) ensure accessibility for the entire population of users.