22484 research outputs found
Sort by
Advancing Perioperative Stroke Prevention: Machine Learning-Based Electroencephalography Monitoring for Detecting Cerebral Ischemia During Carotid Endarterectomy
No full textThis research investigates machine learning (ML) to improve the detection of cerebral ischemia during carotid endarterectomy (CEA) procedures, which carry a higher stroke risk. This work aims to advance perioperative stroke prevention, a severe complication posing difficulty in timely diagnosis. Studies show over 80% of patients experiencing perioperative stroke suffer from significant disability or death. Intraoperative neuromonitoring (IONM) can be done using electroencephalography (EEG) to detect ischemic brain changes that may indicate stroke onset. Manual EEG monitoring poses challenges due to subjective interpretation, potential errors, and limited expert availability.
This work explores integrating ML with EEG to overcome these limitations, pioneering the use of intraoperative EEG for real-time ischemia detection. This contrasts with previous studies that used ML to identify post-stroke EEG patterns. The thesis details the development of ML models to detect ischemic patterns using quantitative EEG (qEEG) features and evaluates them through traditional and time series-aware metrics. Their performance is validated against retrospective labels from a panel of five neurophysiologists.
The results demonstrate that tree-based models effectively distinguish ischemic changes using qEEG. The models’ feature importance ranking reflects the clinical understanding of how ischemia presents in EEG. Time series-aware metrics reveal variations in the models’ abilities to predict the timing and duration of ischemia, essential factors for real-time surgical contexts. Importantly, the ML models achieved non-inferior performance compared to the neurophysiologists’ retrospective review. Of note, moderate agreement between the retrospectively refined labels of five neurophysiologists and minimal agreement with the initial intraoperative labels was found. This emphasizes the inherent subjectivity in EEG interpretation and raises concerns about the reliability of manual IONM.
The findings validate the efficacy of these ML models for intraoperative ischemia detection and their potential for acceptance due to clinically sound explainability. This research contributes to the field by laying a foundation for innovation in perioperative stroke prevention by merging ML approaches with established neuromonitoring techniques to improve surgical outcomes and patient safety
STEM Identity Development: Female Scientists and Mentorship for Middle School Aged Black Girls in the Chicagoland Area
No full textMiddle school Black girls in the Chicagoland area encounter challenges when engaging in Science Technology Engineering and Math (STEM) learning and developing individual STEM
identities. Black girls with good grades in STEM courses are reluctant to pursue STEM careers because of the negative stereotypes they face in school (King & Pringle, 2019),(Lane & Id-Deen,2020). My problem of practice focused on the challenges these girls may encounter when developing individual STEM identities. My dissertation in practice research aimed to reduce the barriers that Black middle school girls encounter when developing individual STEM identities. This inquiry pilot program, known as the Girls Invent Mentorship, engaged middle school girls in partnership with University of Chicago female science mentors (graduate and undergraduate students). Girls (mentees) and female science mentors used an inquiry-based, hands-on science
curriculum developed by the Illinois Student Invention Convention (ISIC) organization.
This mixed method research (Mertens, 2010) project collected data in the form of photos, written science content produced by the girls, science mentors’ reflections, and documented
observations taken by me in my role as the program manager. The research data and observations focused on three girls and their STEM identity development while mentors reflected on their STEM career paths. Each case study highlights the journey of a girl as they develop STEM identities throughout this research study. Data collected revealed trends which included girls having STEM identities along a continuum, strong STEAM identities, and increasing science skills across learning spaces (formal and informal). Girls who engaged with female scientists at the university for several Saturdays showed increased interest in STEM careers and STEM
learning. This pilot program provided insight into the development of a framework that could be used for future STEM mentorship programs hosting middle school girls. In conclusion, this study explored ways in which a mentorship program could support the development of positive STEM
identities and foster an increased interest in STEM careers in Chicago area middle school Black girls
Personalized Affect Assessment
No full textInterest in personalized assessment techniques has grown in recent years in response to the recognition that diagnostic categories fail to capture clinically meaningful heterogeneity. Personalized assessment approaches involve gathering data about psychopathological processes in participants’ everyday lives via digital surveys. Because of their intrusive nature, personalized assessments involve a tradeoff between comprehensiveness—fully capturing relevant psychopathological processes—and brevity—respecting participants’ time. One approach to finding this balance is to develop person-specific assessment instruments that are tailored to be concise and precise. The current study was designed to test the psychometric performance of personalized and non-personalized, ultra-brief assessments of negative affect (NA). P-technique factor analyses were used to create personalized, 3-item sets from the original 15 NA items. Two 3-item strategies, one rationally chosen and the other randomly generated, were used as non-personalized comparisons. A secondary analysis was performed on data from a sub-sample of the Pittsburgh Girls Study (N = 123), in which participants were selected for recent aggressive behaviors. Participants completed 3-week ambulatory assessment protocol after completing clinical interview and baseline assessment. Results demonstrated that all of the ultra-brief NA assessments, both personalized and non-personalized strategies, were relatively successful in replicating the psychometric performance of the full 15 NA item set. There was no clear advantage to using a personalized approach over the non-personalized approaches. These findings suggest that complex, idiosyncratic affective experiences can be assessed idiographically whilst retaining nomothetic value. They further suggest that a standardized, ultra-brief scale will likely be sufficient for capturing NA in research and clinical settings
Fault Mitigation for Spiking-Neural-Network Classification of Neuromorphic Event Streams with Radiation-Induced Noise
No full textThe pursuit of autonomy for on-orbit processing drives the evolution of deep-learning (DL) systems for space applications. Hindered by limited compute capabilities on constrained embedded platforms, alternative technologies are being explored. Amidst the evolution of these DL systems, spiking neural networks (SNNs) coupled with neuromorphic sensing inputs offer an advantage compared to traditional DL frameworks, as SNNs offer lower latency, memory, computation requirements, and subsequent lower power usage. These benefits of neuromorphic systems make them highly amenable to space applications. Given the limited space-flight heritage of neuromorphic systems, the characterization of radiation effects on such algorithms and sensors is valuable for reliable implementations in space. Specifically, characterization of the adverse effects of radiation on neuromorphic systems can lead to development of effective fault-mitigation strategies. This research therefore aims to characterize the effects of simulated, radiation-induced noise on the classification capability of SNNs and explore contemporary fault-mitigation tactics. A fault-injection tool modeled from the results of a neutron-radiation test is leveraged to simulate noise. Using this tool, noise is injected into event-based sensor data and subsequently passed to an SNN for classification. This simulated neuromorphic system provides insight into the radiation-resiliency of SNNs as network hyperparameters such as width and depth are varied, as well as identifies the limitations of current denoising methodology. A novel event-rate denoising method is also introduced to increase the performance of SNN classification under higher magnitudes of the noise, outperforming the standard spatiotemporal filters currently used with event data suitable for radiation environments. The proposed filtering methodology outperforms standard spatio-temporal filtering methodology by at high simulated noise rates (100 events per second) on the complex DVS Gesture dataset, with an improved performance for a wider range of noise rates. The intrinsic reliability of SNNs, coupled with the enhanced event-rate filtering methodology offers a formidable DL alternative
Maroons on Providencia: An Archaeological Look at Self-Emancipation and the Peripheries of Empire in the Western Caribbean
No full textOne powerful form of resistance to colonial powers during the 16th-19th centuries was the formation of Maroon communities. These Maroon communities are primarily comprised of African diasporic peoples who escaped enslavement to live independently from colonial societies. Maroons have existed everywhere that enslaved people existed, including the small, western Caribbean island of Providencia, Colombia. Providencia lies about 150 miles off the coast of Nicaragua and was home to Maroons for much of the colonial period. With no known permanent pre-colonial occupants, the island was colonized by the British and then the Spanish before being officially “abandoned” in 1671. Oral histories however, which are taken in this research as a line of evidence, assert that the island was continuously occupied by Maroon communities formed shortly after the first enslaved Africans were brought to the island.
The research presented here consisted of surface collections and excavations at four Maroon sites on Providencia – Maroon Hill (Site 7), Shortcut (Site 14), and two sites at Manchaneel Hill (Sites 20 and 21). Archaeological evidence suggests a different Maroon lifeway and trajectory on Providencia than is found in many other archaeologically studied Maroons. There is no break in the archaeological record suggesting a “period of abandonment” – instead this period should be known as the “Maroon period” as Maroons were the only permanent residents on the island. During this time, archaeological evidence suggests that Maroons were trading extensively with outsiders for European materials. In addition to this, this research presents the first evidence of colonoware on the island.
Community archaeology with the Native Raizal and a framework of Black resistance serve as essential research paradigms. Studying the process by which Maroons survived, adapted, and thrived in various settings across the African diaspora is essential to a more holistic understanding of colonial history. The Maroons on Providencia created their own culture and identity by navigating a growing global system through illicit trade as well as relatively autonomous living far from any formal colonial power, challenging current assumptions about what it means to be Maroon
Novel AMTB chimeric analogs as antagonists of the cold menthol receptor TRPM8
No full textTRPM8, also known as the cold menthol receptor, is a calcium-permeable ion channel
of the transient receptor potential (TRP) channel superfamily that is second only to voltage-gated
potassium channels. TRPM8 knockout mice show no reaction to decreases in environmental
temperature and exhibit minimal sensitivity to cold stimuli after traumatic nerve injury. Existing
TRPM8 antagonists can alleviate pain but disrupt core body temperature regulation. We
hypothesize that the activation of human TRPM8 by ligands and environmental cold occurs in a
different part of the channel than the region responsible for TRPM8 activation involved in
thermoregulation, such as maintaining core body temperature or responding to hypothermia. Our
approach is to generate novel, mode-selective antagonists capable of performing strong, anchoring
interactions in the common binding site sensitive to chemical and environmental cold stimuli, with
chimeric molecules of the TRPM8 antagonist AMTB and our existing VBJ series, to reverse cold
hypersensitivity in the absence of a thermoregulatory effect. Here, we report a series of AMTB
VBJ chimeric analogs that inhibit chemical (icilin)-induced responses at human TRPM8 in a dose
dependent manner. Compounds 6a, 6e, and 6i inhibit icilin-stimulated responses with IC50 values
of 4.4 ± 0.8 µM, 3.3 ±1.8 µM, and 2.6 ± 2.1 µM, respectively. Compounds 6a and 6e possess
optimum logD values and in-vitro T1/2 values of 18.2 and 23.3 min in mouse liver microsomes
Genetic Investigations of Cleft Lip with or without Cleft Palate: A Literature Review and Genome-wide Association Study
No full textCleft lip with or without cleft palate (CL/P) is one of the most common congenital anomalies worldwide, with a complex etiology involving both genetic and environmental fac- tors. This thesis provides a comprehensive literature review of the epidemiology, psychology, healthcare costs, and genetics of CL/P, followed by an original genome-wide association study (GWAS) and downstream analyses in a Filipino population. The literature review traces the historical understanding of CL/P, discussing its epidemiology across different populations, the psychological impact on affected individuals, and the substantial healthcare costs as- sociated with treatment. The review then discusses the genetic aspects of CL/P, covering heritability estimates, segregation analyses, linkage studies, genome-wide association stud- ies (GWAS), gene-environment interactions, sequencing studies, animal models, and phe- notypic heterogeneity. The original research presented is a GWAS of CL/P in a Filipino cohort of 1,399 unrelated individuals (882 cases, 517 controls). The study identified a novel locus, KCNQ5, and replicated associations at three known loci: ARHGAP29, IRF6, and SHROOM3. A gene-based test using mBAT-combo highlighted potential masking effects in several genes, including IRF6. A transcriptome-wide association study (TWAS) further sup- ported the involvement of known clefting loci. Fine-mapping analyses and conditional and joint multiple-SNP analysis (COJO) provided insights into putative causal variants, partic- ularly the missense variant rs2235371 in IRF6, which showed strong functional evidence and high linkage disequilibrium with the top signal. This thesis underscores the importance of conducting genetic studies in diverse populations to better understand the genetic architec- ture of complex traits like CL/P and to address disparities in genetic research. The public health relevance of this work lies in its potential to improve our understanding of CL/P eti- ology, which could lead to better prevention strategies, earlier diagnosis, and more targeted treatments, ultimately reducing the burden of this condition on affected individuals, families, and healthcare systems. The findings contribute to the growing body of knowledge on the genetic basis of CL/P and provide a foundation for future research aimed at elucidating the etiology of this condition and improving strategies for prevention, diagnosis, and treatment
Teaching and Learning Through Dialog: Oral History Interviewing Methods as Experiential Learning
No full textHow can educators help students to embrace new ways of knowing, particularly ways of knowing that interrogate or disrupt traditional notions of authority and objectivity? This dissertation explores the potential of oral history as a pedagogical tool to challenge dominant epistemologies and empower students. By intertwining composition studies and oral history, I argue that centering the experiential knowledge of everyday people is a liberatory teaching practice that can cultivate critical thinking skills.
Through a reflexive examination of my teaching experiences in high school and university classrooms, I demonstrate how oral history can be integrated into diverse learning environments. By engaging students in storytelling and interviewing activities, I sought to create spaces for students to develop their voices, critique hegemonic power structures, and co-construct knowledge. This research contributes to the fields of composition studies and oral history by offering practical examples of how these methodologies can be combined to create transformative learning experiences
Duties in Retrospect: Ethical Limitations to the Psychiatrist’s Duty to Warn
No full textTarasoff v. Regents of the University of California established that mental health professionals in California have a legal duty to seek to protect individuals who are foreseeably endangered by violent patients, including warning them of threats of harm. In the wake of this landmark case, many other states mandated legal duties to protect in general, or warn specifically, either through statute or case law. At the same time, an ethical discourse emerged, challenging the duty to warn on the grounds that it forces physicians to breach confidentiality, an ethical obligation fundamental to medical practice. This thesis examines the origins of the legal duties to protect and warn, the consequences of these duties to the practice of psychiatry, and the conflict between legal and ethical obligations to patients and the public. Although psychiatrists have a professional obligation to maintain patient confidentiality, this obligation may be set aside in the interest of public safety. However, limitations in predicting violence thwart the psychiatrist’s ability to recognize when confidentiality should be set aside. Ultimately, imposing a mandatory legal duty to protect, and specifically a duty to warn ignores the complexity surrounding patient violence. Permissive laws are more sensible, but even these may not prevent violence, making them less justifiable ethically
Quantifying SARS-CoV-2 Infection Risks in Inpatient Psychiatric Settings: A Risk Assessment Model
No full textBackground/Objective: In March, 2020, SARS-CoV-2 precipitated a global pandemic. Highly infectious and more likely to cause severe disease in those with underlying comorbidities, the virus necessitated surveillance and risk mitigation strategies. While vaccination efforts have seen relaxing of many of these measures, certain populations, namely the inpatient psychiatric population, remain at-risk for infection and complications.
Methods: This retrospective cohort study identified significant demographic, health related, and facility-based predictors of hospital acquired COVID-19 infection. Relevant data was collected from admissions and readmissions at an inpatient psychiatric hospital from June 2020 to June 2022. Univariate Chi-squared and Student’s t-tests were employed to identify significant differences in frequencies of predictors between nosocomial COVID-19 cases and controls; covariates reaching significance were included in a Cox Proportional Hazards model to assess temporal infection risks. A sensitivity analysis was conducted for Cox regression excluding positive SARS-CoV-2 tests ambiguous with respect to nosocomial origin.
Results: Of 10,845 admissions, 40 tested positive for SARS-CoV-2 during hospital stay. Patients infected with SARS-CoV-2 were older, (p<0.001), more likely to be discharged from certain units (p<0.001), less likely to be vaccinated (p<0.001), showed higher prevalence of cardiovascular disease (p<0.001) and had longer lengths of stay (p=0.0019). Cox regression identified two discharge units, a general inpatient (HR: 57.2; 95% CI: 5.870, 557.320; p=0.0005) and intensive care unit (HR: 32.4; 95% CI: 3.239, 324.194; p=0.0031), as significantly associated with increased risk of infection. Sensitivity analysis excluding cases of unconfirmed nosocomial origin yielded no significant predictors of SARS-CoV-2 infection risk.
Conclusion: In an inpatient psychiatric setting, unit of discharge, cardiovascular disease, age, vaccination status and length of stay were associated with nosocomial SARS-CoV-2 infection risk. These factors are known correlates with diagnosis, severity and duration of psychiatric illness. Thus, risk reduction efforts, especially vaccination, should be targeted to patients with severe mental illness. By assessing predictors of hospital-acquired SARS-CoV-2 infection and identifying areas of focus for interventions in inpatient psychiatric settings, this study adds to important public health literature on the inpatient psychiatric population