DSpace@RPI (Rensselaer Polytechnic Institute)
Not a member yet
6809 research outputs found
Sort by
The development of the Human Health Exposure Analysis Resource (HHEAR) Data Repository for environmental epidemiology research
No full textImplementation of the exposome paradigm is a critical aspect of the next generation of environmental health research studies. To spur exposomics research, the U.S.-based Human Health Exposure Analysis Resource (HHEAR) provided scientific investigators access to both laboratory and statistical analyses aimed at incorporating and expanding the breadth of biological markers of environmental exposures within their research. To extend the benefits of this program to the broader scientific community, the HHEAR Data Center established a public data repository to facilitate pooling and sharing of data generated by the HHEAR program. All HHEAR investigators deposited epidemiologic data on study participants, to accompany the biomarkers of exposure generated by the HHEAR laboratories. The latest semantic technologies are used to efficiently conduct data standardization across studies and promote data sharing by aligning the repository with the FAIR (Findable, Accessible, Interoperable, Reusable) data principles. This includes standardizing individual study data to a common ontology and representing data within a knowledge graph. A clear user interface enables search, construction, and download of customized datasets and maintenance of provenance through use of digital object identifiers. The repository will eventually contain information from 35,989 individuals across 55 environmental health studies, including data on biomarkers of environmental exposures, sociodemographics, health outcomes, and physical and mental assessments. All data are freely downloadable for reuse after a brief application for data access. Designed to support cutting-edge research and education, the HHEAR Data Repository provides a rich, harmonized resource of exposure biomarkers and corresponding health data from diverse study populations
The effect of anterior plate stiffness on load-sharing and adjacent level biomechanics following arthrodesis of the cervical spine
No full textDecember 2024School of EngineeringLow back and neck pain are estimated to affect approximately 80% of individuals at least once in their lifetime. One common cause underlying both back and neck pain is degenerative disc disease (DDD). This disease is characterized by an overall loss in intervertebral disc height causing subsequent pain. When pain becomes chronic the current gold standard for pain relief in the cervical spine is an anterior discectomy and fusion (ACDF). This surgical intervention involves restoring lost disc height by inserting an anterior cervical cage in the intervertebral disc space and using an anterior cervical plate for stability during healing. Successful fusion occurs when a bony bridge forms between two adjacent vertebrae, relieving pain. While this surgery has a high success rate in one-level fusions, a common pathology that can occur is adjacent-level degeneration following the ACDF procedure. This degeneration occurs at the super and sub adjacent levels to the index level. There is no consensus whether this accelerated degeneration is caused by biomechanical changes or if an individual is pre-disposed to degenerative changes as a natural progression of DDD, independent of surgical intervention.Prior research suggests that anterior cervical plate stiffness could affect the biomechanical environment following instrumentation in an ACDF by altering load sharing through the intervertebral disc space and a shift in the instantaneous axis of rotation (IAR) location. Therefore, to study biomechanical changes at the adjacent levels in vitro we developed a range of novel low stiffness anterior cervical plates and a low stiffness cage using CAD and finite element analysis (FEA). The anterior cervical cage was instrumented with strain gauges to develop a “smart” implant to monitor interbody loads in real-time. A novel mechanical testing system was also designed to apply physiologically relevant loading conditions in vitro to more accurately study the changes to the biomechanical environment at adjacent levels.
We hypothesized that there would be a correlation between anterior cervical plate stiffness and load-sharing: as implant stiffness decreases load sharing between the plate and cage increases. We also hypothesized that there would be a correlation between implant stiffness and intersegmental motion, facet joint loads, and the location of the instantaneous axis of rotation: as stiffness decreases there will be (a) decreased intersegmental motion at the super-adjacent and sub-adjacent motion segments, (b) decreased loading on the facet joints of the index level and super- and sub-adjacent levels, and (c) a more physiologic instantaneous axis of rotation when compared to stiffer cervical plates.
Results demonstrated that anterior cervical plate stiffness does dictate the biomechanical environment by altering load-sharing and the location of the IAR. We further demonstrated in vitro that anterior cervical plate stiffness was one factor that could affect adjacent level biomechanics leading to degenerative changes and should be considered when studying adjacent level disease. However, underlying pathological changes also appeared to play a role in the biomechanical environment both at the index and adjacent levels.Ph
Live coding the collective: reimagining experimental music practices through technics of decentralization
No full textDecember 2024School of Humanities, Arts, and Social SciencesThis dissertation explores how decentralization technologies could transform experimental music practices, particularly computer-based musicking such as live coding and electroacoustic music composition. Grounded in Gilbert Simondon's relational ontology, the research first establishes an innovative theoretical framework that identifies and analyzes four affirmationist subjective modes in contemporary technoculture: tachyphilia (attraction to speed), thermophilia (affinity for heat), hydrophilia (liquid-driven dissolution), and atmophilia (vapor-seeking diffusion). These modes trace an intensificatory trajectory from industrial acceleration through nuclear and electromechanical calefaction, to digital liquefaction and AI-driven evaporation, providing a novel lens for understanding technological and aesthetic developments. This imaginative framework illuminates a speculative interplay between key technologies of our time, particularly the relationship between blockchain's distributed liquidity and AI's ubiquitous imaginality. The study employs a practice-based research methodology combining artistic production, software development, and theoretical synthesis. Two major practical experiments are presented: a blockchain-based system for collaborative electroacoustic music composition incorporating a novel Proof of Creative Contribution (PoCC) consensus protocol, and an extended framework for live coding based on SCTweets (code snippets of algorithmic music shared through social media) that emphasizes modulatory and transductive playing techniques. These experiments are complemented by the creation of an AI-generated film adaptation of Nick Land's essay 'Meltdown,' revealing serendipitous moments of poetic signification that emerge from the interplay between philosophical text, AI generation, and human curation and editing while exploring the aesthetic dimensions of accelerationist theory. Technical contributions include the development of open-source libraries for extending live coding capabilities through feedback networks and adaptive feature mapping, alongside novel implementations of multimodal interfaces incorporating MIDI control, gesture recognition, and biosensing. While the dissertation argues that blockchain technology offers concrete possibilities for realizing mutualist, collaborative, and commons-oriented music-making practices, it acknowledges that these technologies must be critically examined and thoughtfully implemented to avoid reinforcing existing power structures and inequalities. The research proposes organizing experimental music practices as Decentralized Autonomous Organizations (DAOs), suggesting new models for community governance and value creation within artistic communities, while speculatively exploring how live coding principles might transform blockchain operations themselves through real-time, collaborative programming of decentralized protocols. This investigation advances both theoretical understanding and practical applications in the emerging field of decentralized music systems, contributing to ongoing discussions about technology's role in artistic practice while offering specific frameworks and tools for implementing more collaborative and equitable forms of music creation. The research concludes by proposing a rhizomatic ecosystemization of experimental music practices, envisioning an organic network of creative exchange between distinct musical communities mediated by commons-based Distributed Ledger Technology infrastructures.Ph
Neurosurgical medical image processing & innovations for imaging intracranial tumors
No full textJuly 2024School of EngineeringMeningiomas make up about 20% of all intracranial tumors in men and 38% in women. Its prevalence is estimated at 97.5/100,000 with 2 and 5 year survival rates at 81% and 69%. Meningiomas make up 37.6% of all primary CNS tumors and 53.3% of all benign CNS tumors. Meningiomas are also the most common central nervous system tumors. The majority of meningiomas are benign in nature, their slow growth results in very gradual symptoms that remain unnoticed until tumors grow much larger in size . These usually come to the attention of patients when they experience cranial neuropathies (i.e. vision loss), seizures and hemiparesis which can sometimes stem from the pressure being exerted on regions of the brain. Diagnostic imaging of meningiomas is conducted through MR imaging before embolization and resection. Additional MR and histological imaging is conducted after resection. Surgical resection remains the golden clinical standard for treating large, symptomatic, and highly vascularized meningiomas, with radiosurgery used as an alternative for small, recurrent, or residual tumors. Preoperative embolization is a neoadjuvant therapy that constitutes a minimally invasive procedure in which embolic agents are delivered via a catheter to occlude arterial supply to tumors. The rationale for preoperative embolization has traditionally been to minimize intraoperative blood loss during surgical resection of highly vascular tumors. Systematic reviews have demonstrated that intraoperative blood loss is an unreliable metric as access to tumors and inaccurate measurements make this an erroneous measure. There is currently no tool to intraoperatively quantify the percent of the tumor that has been embolized, forcing surgeons to visually assess angiography data to determine success. This subjective endpoint results in poor reproducibility and limits the ability to automate future tools to optimize the technique. There is a clear need for the creation of an intraoperative image-processing tool to rapidly and accurately assess tumor embolization percentages. This is a report of work that aims to introduce quantitative image analysis tools to assess the extensiveness of endovascular embolization of meningiomas as well as the direct physiological effects. These image analysis tools applied in: MR data, angiography data, and histology data.DEn
Hydrologic and spatiotemporal controls on pyrogenic carbon export from fire-affected, coastal california streams
No full textMay2025School of ScienceWildfires play a crucial role in shaping ecosystems but are increasingly driven by anthropogenic factors, leading to shifts in fire regimes and their associated environmental impacts. One significant consequence of fire is the production and mobilization of pyrogenic carbon (PyC), a complex mixture of thermally altered organic compounds that influence carbon cycling in terrestrial and aquatic systems. Recent increases in wildfire frequency and severity, driven by climate change and land-use alterations, have amplified the need to understand the environmental fate of fire-derived materials in order to mitigate and prepare for continued fire perturbations. The chapters of this dissertation aim to improve our understanding of PyC in fire-affected aquatic systems by assessing the efficacy of current PyC quantification methods, determining the variability in PyC concentration and character, and/or identifying any drivers of PyC export from riverine systems. In Chapter 1, I introduce the two specific fractions of PyC targeted in following chapters: levoglucosan (a rapidly degraded anhydrosugar) and black carbon (a refractory condensed aromatic compound class). In Chapter 2, I assess the degree of impact free benzenepolycarboxylic acids (BPCAs) have on the current dissolved black carbon quantification method. Black carbon is most commonly quantified via oxidation of its characteristic condensed aromatic structures into BPCA molecular markers for analysis. Results show that dissolved black carbon concentrations are not artificially overestimated by the presence of free BPCAs as long as only certain conversion factors are utilized. In Chapter 3, I investigate two paired temporary streams that were equivalently impacted by the 2020 SCU Lightning Complex Fires. Through high-temporal resolution measurements, watershed-specific drivers of dissolved organic and black carbon export were identified. These streams also showcased that while there is a significant first-flush of dissolved black carbon exported during the first stream wet-up post-fire, the greatest concentrations were observed during wet-up of the second year post-fire. In Chapters 4 and 5, I investigate the post-fire PyC export behavior from five coastal mountain watersheds variably impacted by the 2020 CZU Lightning Complex Fires. In Chapter 4, results from black carbon concentrations suggest that hydrology rather than the percentage of watershed area burned is the primary driver of post-fire export behavior. In Chapter 5, levoglucosan concentrations were compared to that of black carbon from these same watersheds. Distinct molecular class-specific characteristics were revealed. Levoglucosan, being highly soluble, was rapidly mobilized during the first events post-fire. In contrast, black carbon concentrations were more closely linked with discharge and the ratio of dissolved black carbon to organic carbon increased with time since fire highlighting the role of black carbon aging increasing solubility and therefore mobilization. In Chapter 6, I contextualize the findings from the previous chapters, commenting on spatiotemporal controls on PyC export and make suggestions for future avenues of research. Overall, these findings contribute to a more comprehensive understanding of PyC fluxes and their implications for carbon budgets, water quality, and role in the Earth system.Ph
Exploring efficacy in digital therapeutics: serious games for theory of mind training and visual rehabilitation
No full textMay2025As digital media increasingly transforms healthcare and education, the unique affordances of serious video games present distinct challenges and opportunities, setting them apart from traditional serious games. While conventional serious games often superficially gamify clinical practices, diminishing meaningful engagement, or replicate clinical protocols so rigidly that intrinsic player motivation suffers, serious video games uniquely offer interactive affordances that could effectively reconcile clinical precision with authentic player engagement. This dissertation examines this critical tension through the iterative development and empirical evaluation of two digital therapeutic prototypes: \textit{Emotion Adventure}, designed to foster Theory of Mind, the ability to understand and interpret others’ emotional and mental states, in children with Autism Spectrum Disorder, and \textit{Eye Rehab}, a virtual reality game aimed at improving stroke-related visual impairments. These prototypes were systematically designed and evaluated using the Mechanics, Dynamics, and Aesthetics (MDA) framework, which methodically connects foundational game mechanics, emergent player dynamics, and experiential aesthetics to ensure balanced game design. In usability evaluations, \textit{Emotion Adventure} employed a narrative-driven approach to successfully promote empathic decision-making and maintain player engagement within structured gameplay interactions. However, given the complexities and resource demands involved in empirically measuring cognitive therapeutic outcomes, a second prototype, \textit{Eye Rehab}, was developed. This physiological digital therapeutic utilized virtual reality-based gaze interactions to precisely target measurable improvements in visual alignment and ocular motor functions, validated clinically through the Lancaster Red-Green test, a standardized diagnostic tool used to assess ocular alignment and muscle function. Building upon insights gained through these prototypes, this dissertation hypothesizes a replicable design approach termed \textit{Selective Simulation}, which strategically embeds essential therapeutic actions directly into core gameplay mechanics. Unlike earlier theoretical concepts such as persuasive or applied games that offer generalized guidance, \textit{Selective Simulation} provides concrete and empirically informed design principles to intentionally integrate therapeutic activities within engaging game mechanics. Ultimately, this dissertation contributes to the broader field by proposing a replicable and structured framework for serious video game design, bridging theoretical insights from media studies, cognitive psychology, and human-computer interaction with methodologies rooted in clinical practice. This interdisciplinary approach underscores the distinct potential and complexity of video games as digital therapeutics, advocating for designs that rigorously balance therapeutic efficacy and engaging gameplay.Ph
Deep probabilistic and generative models for x-ray based imaging and ecg
No full textMay2025School of EngineeringArtificial intelligence (AI) is poised to transform modern medicine and, in particular, medical imaging, as the need for healthcare services continues to outstrip available resources worldwide. Deep learning models show promise in enhancing clinical workflows and patient outcomes through data-driven diagnoses and improved image quality. Despite the pressing need and considerable research efforts, several challenges have delayed the integration of AI into healthcare: Neural networks often lack explainability, making overconfident inferences on out-of-sample data, and are susceptible to adversarial attacks—small, targeted input perturbations capable of fooling otherwise accurate networks. Furthermore, large models are data-hungry, yet patient images and information are legally guarded by health privacy protections. Future virtual clinical trials for medical device validation also require high-quality synthetic datasets that sufficiently represent rare pathologies and population demographics. While deep generative models may address these data gaps, their outputs often lack clinical fidelity despite appearing visually convincing. Finally, many image-enhancement tasks, such as deblurring, lack the ground-truth labels necessary for supervised learning, forcing reliance on simulated degradations that can introduce artifacts when applied to real-world data. Compounding the above issues is the high-dimensional nature of most medical signals, invalidating solutions that cannot be feasibly scaled. The following dissertation investigates solutions to several of the aforementioned challenges within specific applications, primarily leveraging deep probabilistic and generative models. First, we examine how diverse deep ensembles, aided by a feature decorrelation mechanism, can improve adversarial robustness in high-dimensional tasks like electrocardiogram classification. Next, drawing inspired by the 2023 AAPM Deep Generative Modeling Challenge, we employ denoising diffusion probabilistic models to produce synthetic medical images that are realistic in both visual appearance and clinical relevance—an effort that earned first place in the competition. Finally, we adapt state-of-the-art simulation techniques to create realistic, system specific degradations to train deep deblurring models for photon-counting computed tomography. By scaling a diffusion model to 3D through a joint 2D inference process and disentangling noise from signal prior to deblurring, we successfully mitigate texture distortions and improve performance on real-world data.Ph
Towards scalable and generalizable multi-objective learning, provably
No full textMay2025School of EngineeringMany learning tasks in today's real-world systems inherently involve multiple objectives. In such problems, objectives must balance multiple performance metrics, including fairness, safety, privacy, and accuracy, or address potentially conflicting objectives from different entities that are optimized jointly to facilitate data and knowledge sharing. A fundamental question here is how to handle multiple objectives in a principled manner, such that it allows pursuing individual objectives but also preserves the benefits of collaborative learning. In this context, the thesis will put forth a new grounded framework to address multi-objective learning from three complementary aspects: modeling, optimization, and generalization. The research will enhance multi-objective knowledge extraction with theoretical guarantees (optimization, generalization, conflict mitigation, and their trade-offs) in future systems. Dedicated to addressing these challenges, this thesis can be summarized as follows: In the first part, we study the multi-objective optimization algorithms. 1) We revisit some stochastic variants of the multi-gradient descent algorithm (MGDA) in the unconstrained setting, propose a new variant and a framework to analyze the optimization error and the conflict avoidance. Then we apply the framework to existing stochastic algorithms, which leads to improved analysis. 2) When there are pre-specified preferences over objectives, we consider a formulation through constrained vector optimization, where the preferences are modelled through a cone-induced partial order and the constraint functions. 3) To prioritize optimality over preference satisfaction, we consider another formulation through optimization on the Pareto set. Under the formulations of 2) and 3), efficient gradient-based algorithms are developed with convergence rate guarantees. In the second part, we study the generalization of multi-objective learning. 1) In the unconstrained setting, we provide an algorithm-dependent generalization bound for the stochastic variants of MGDA. Furthermore, combined with the optimization error and conflict avoidance analysis, it forms a unified framework to analyze the three errors and their trade-offs. 2) In the meta learning problem, which is a special example of multi-objective learning, we analyze the modeling and generalization errors in the mixed linear regression setting. The results suggest that model adaptation helps reduce modeling error, increasing the number of tasks or objectives helps reduce generalization error, and uncertainty modeling through Bayesian inference further helps reduce generalization error.Ph
The Logic of Bias: Using Cognitive Architecture to Explore Interactions Between Cognitive Abilities and Decision Error
No full textThe traditional view of biases being cognitive imperfection has been challenged by several strains of research, such as the PSI cognitive architecture. Here, biases are considered to be engineered by evolution, to prevent dissatisfaction and assist subsequent satisfaction of human needs. PSI's general assumption of higher skills and reasoning capacities alleviating biases has been recently called into question, as high numeracy was associated with an exacerbated effect of political bias. We conduct two studies, the results of which indicate that the basis for this effect 1) does not represent a general cognitive fallacy caused by modulations of perceptional and attentional processes, 2) nor is rooted in the long-term forming of habituated action patterns, associated with prior beliefs. This strengthens the evidence for it to be specific to group dynamics with strong affiliative bounds. Further, we propose a set of revisions to PSI, necessary to model this expert bias phenomenon
Decision models for volunteer management in nonprofits: enhancing engagement, satisfaction, and retention
No full textAugust2025School of EngineeringNonprofit organizations (NPOs) rely heavily on volunteers to support community needs and provide a variety of essential services. However, they face challenges in effectively utilizing limited volunteer and employee resources while navigating uncertainties in volunteer motivations, task preferences, and participation behaviors. A central challenge is making volunteer-to-task assignment decisions that can accommodate the needs of all three stakeholders (nonprofit, community members, volunteers), in environments representing different volunteer behaviors. The difficulty in quantifying volunteer motivations and satisfaction further complicates these decisions, making it challenging to align organizational and community benefits with volunteer retention. To address these complexities, this dissertation develops multiple optimization approaches that enhance volunteer management strategies while assessing their impact through computational studies capturing diverse volunteer populations with varying preferences and arrival patterns. The first study presents a deterministic multi-period integer linear program for optimizing task assignments for volunteers and NPO employees, balancing factors such as task urgency, volunteer training, and task or location preferences. Using a food bank case study, the model incorporates uncertainties in volunteer task preferences and retention. It highlights the trade-offs between operational performance and the impact of current assignments on future volunteer retention. Simulation experiments examine factors like volunteer availability, preferences, and retention uncertainties. Results demonstrate that the proposed model significantly improves key performance indicators compared to a benchmark policy that ignores retention considerations. The proposed approach is found to be beneficial to NPOs, even when an NPO has uncertainty in their estimates for volunteer preferences and volunteer retention threshold values.
The second study explores how integrating characteristics of both volunteers and NPO tasks can enhance strategic volunteer-to-task assignments and improve volunteer satisfaction. Drawing on insights from qualitative literature, the study incorporates key motivators—such as opportunities for skill development, contributing to a meaningful cause, collaboration, and networking—into a static integer programming framework.
A key challenge in NPOs is balancing volunteer autonomy with effective task assignments, as existing methodologies rarely incorporate volunteer-driven task selection. To address this, the study develops a personalized task recommendation system (Menu Creation Integer Program), which generates personalized task menus (i.e., a subset of tasks) for each volunteer based on demographic information and preferences collected during onboarding. A multi-scenario approach simulates volunteer selections across multiple iterations, and a consensus algorithm refines the final menus to be presented to volunteers to encourage task selection. At this stage, volunteer’s task selections (willingness) are collected. The study then implements the Group Creation Integer Program, which forms homogeneous volunteer groups by considering skill alignment and interpersonal affinity.
Existing research often overlooks structured group formation for complex NPO tasks, focusing instead on one-time assignments. By developing a systematic approach to strategic group creation, this study bridges a critical gap in volunteer management. Given the combinatorial complexity of group formation and the inherent uncertainty in volunteer availability, the framework prioritizes both organizational efficiency and volunteer experience. Empirical data from a partner NPO informs the design of experiments, demonstrating that personalized task recommendations and strategic group formation significantly enhance both volunteer satisfaction and NPO effectiveness, even when volunteers are highly selective. A case study on a remote NPO with online volunteers further illustrates the benefits and trade-offs of incorporating structured volunteer groups into assignment strategies. The findings highlight the trade-offs between organizational goals and volunteer preferences, offering a scalable approach to optimizing volunteer management in NPOs.
The third study extends the static methodology from Study 2 by developing a dynamic rolling horizon model to assess the long-term impact of group-task assignments on volunteer retention. This model incorporates past volunteer assignments and group compositions to inform strategic assignments over multiple periods while capturing retention. A scenario-based dynamic approach is used to account for uncertainty in volunteer-to-task assignments, with multiple scenarios and associated probabilities capturing variations in menu offerings and task allocations. Chance constraints are incorporated to ensure optimal assignments remain feasible under uncertainty. The focus of this methodology is to enhance both immediate satisfaction and long-term engagement, ultimately boosting nonprofit sustainability and operational efficiency.
This dissertation advances the field of nonprofit operations management by providing actionable methodologies to enhance task completion, resource utilization, and long-term volunteer engagement through volunteer group creation, helping NPOs sustain their vital role in addressing community needs.Ph