- Browse by Author
Browsing by Author "Jang, Jeong Hoon"
Now showing 1 - 10 of 11
Results Per Page
Sort Options
Item An integrative latent class model of heterogeneous data modalities for diagnosing kidney obstruction(Oxford University Press, 2024) Jang, Jeong Hoon; Chang, Changgee; Manatunga, Amita K.; Taylor, Andrew T.; Long, Qi; Biostatistics and Health Data Science, School of MedicineRadionuclide imaging plays a critical role in the diagnosis and management of kidney obstruction. However, most practicing radiologists in US hospitals have insufficient time and resources to acquire training and experience needed to interpret radionuclide images, leading to increased diagnostic errors. To tackle this problem, Emory University embarked on a study that aims to develop a computer-assisted diagnostic (CAD) tool for kidney obstruction by mining and analyzing patient data comprised of renogram curves, ordinal expert ratings on the obstruction status, pharmacokinetic variables, and demographic information. The major challenges here are the heterogeneity in data modes and the lack of gold standard for determining kidney obstruction. In this article, we develop a statistically principled CAD tool based on an integrative latent class model that leverages heterogeneous data modalities available for each patient to provide accurate prediction of kidney obstruction. Our integrative model consists of three sub-models (multilevel functional latent factor regression model, probit scalar-on-function regression model, and Gaussian mixture model), each of which is tailored to the specific data mode and depends on the unknown obstruction status (latent class). An efficient MCMC algorithm is developed to train the model and predict kidney obstruction with associated uncertainty. Extensive simulations are conducted to evaluate the performance of the proposed method. An application to an Emory renal study demonstrates the usefulness of our model as a CAD tool for kidney obstruction.Item A Bayesian multiple imputation approach to bivariate functional data with missing components(Wiley, 2021) Jang, Jeong Hoon; Manatunga, Amita K.; Chang, Changgee; Long, Qi; Biostatistics and Health Data Science, School of MedicineExisting missing data methods for functional data mainly focus on reconstructing missing measurements along a single function-a univariate functional data setting. Motivated by a renal study, we focus on a bivariate functional data setting, where each sampling unit is a collection of two distinct component functions, one of which may be missing. Specifically, we propose a Bayesian multiple imputation approach based on a bivariate functional latent factor model that exploits the joint changing patterns of the component functions to allow accurate and stable imputation of one component given the other. We further extend the framework to address multilevel bivariate functional data with missing components by modeling and exploiting inter-component and intra-subject correlations. We develop a Gibbs sampling algorithm that simultaneously generates multiple imputations of missing component functions and posterior samples of model parameters. For multilevel bivariate functional data, a partially collapsed Gibbs sampler is implemented to improve computational efficiency. Our simulation study demonstrates that our methods outperform other competing methods for imputing missing components of bivariate functional data under various designs and missingness rates. The motivating renal study aims to investigate the distribution and pharmacokinetic properties of baseline and post-furosemide renogram curves that provide further insights into the underlying mechanism of renal obstruction, with post-furosemide renogram curves missing for some subjects. We apply the proposed methods to impute missing post-furosemide renogram curves and obtain more refined insights.Item Cognitive and Neuronal Link With Inflammation: A Longitudinal Study in People With and Without HIV Infection(Wolters Kluwer, 2020-12) Anderson, Albert M.; Jang, Jeong Hoon; Easley, Kirk A.; Fuchs, Dietmar; Gisslen, Magnus; Zetterberg, Henrik; Blennow, Kaj; Ellis, Ronald J.; Franklin, Donald; Heaton, Robert K.; Grant, Igor; Letendre, Scott L.; Biostatistics, School of Public HealthBackground: Across many settings, lack of virologic control remains common in people with HIV (PWH) because of late presentation and lack of retention in care. This contributes to neuronal damage and neurocognitive impairment, which remains prevalent. More evidence is needed to understand these outcomes in both PWH and people without HIV (PWOH). Methods: We recruited PWH initiating antiretroviral therapy and PWOH at 2 sites in the United States. One hundred eight adults were enrolled (56 PWOH and 52 PWH), most of whom had a second assessment at least 24 weeks later (193 total assessments). Tumor necrosis factor alpha, monocyte chemotactic protein-1 (MCP-1), neopterin, soluble CD14, and neurofilament light chain protein (NFL) were measured in plasma and cerebrospinal fluid (CSF). Using multivariate models including Bayesian model averaging, we analyzed factors associated with global neuropsychological performance (NPT-9) and CSF NFL at baseline and over time. Results: At baseline, higher CSF MCP-1 and plasma sCD14 were associated with worse NPT-9 in PWH, while CSF HIV RNA decrease was the only marker associated with improved NPT-9 over time. Among PWH, higher CSF neopterin was most closely associated with higher NFL. Among PWOH, higher CSF MCP-1 was most closely associated with higher NFL. After antiretroviral therapy initiation, decrease in CSF MCP-1 was most closely associated with NFL decrease. Conclusion: Monocyte-associated CSF biomarkers are highly associated with neuronal damage in both PWH and PWOH. More research is needed to evaluate whether therapies targeting monocyte-associated inflammation may ameliorate HIV-associated neurobehavioral diseases.Item The effect of in-person primary and secondary school instruction on county-level SARS-CoV-2 spread in Indiana(Oxford, 2021) Bosslet, Gabriel T.; Pollak, Micah; Jang, Jeong Hoon; Roll, Rebekah; Sperling, Mark; Khan, Babar; Medicine, School of MedicineBackground To determine the county-level effect of in-person primary and secondary school reopening on daily cases of SARS-CoV-2 in Indiana. Methods This is a county-level population-based study using a panel data regression analysis of the proportion of in-person learning to evaluate an association with community-wide daily new SARS-CoV-2 cases. The study period was July 12-October 6, 2020. We included 73 out of 92 (79.3%) Indiana counties in the analysis, accounting for 85.7% of school corporations and 90.6% of student enrollement statewide. The primary exposure was the proportion of students returning to in-person instruction. The primary outcome was the daily new SARS-CoV-2 cases per 100,000 residents at the county level. Results There is a statistically significant relationship between the proportion of students attending K-12 schools in-person and the county level daily cases of SARS-CoV-2 28 days later. For all ages, the coefficient of interest (β) is estimated at 3.36 (95% CI: 1.91—4.81; p < 0.001). This coefficient represents the effect of a change the proportion of students attending in-person on new daily cases 28 days later. For example, a 10 percentage point increase in K-12 students attending school in-person is associated with a daily increase in SARS-CoV-2 cases in the county equal to 0.336 cases/100,000 residents of all ages. Conclusion In-person primary and secondary school is associated with a statistically significant but proportionally small increase in the spread of SARS-CoV-2 cases.Item The Effect of In-Person Primary and Secondary School Instruction on County-Level Severe Acute Respiratory Syndrome Coronavirus 2 Spread in Indiana(Oxford University Press, 2022-01-07) Bosslet, Gabriel T.; Pollak, Micah; Jang, Jeong Hoon; Roll, Rebekah; Sperling, Mark; Khan, Barbara; Medicine, School of MedicineBackground: Our goal was to determine the county-level effect of in-person primary and secondary school reopening on daily cases of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Indiana. Methods: In this county-level, population-based study, we used a panel data regression analysis of the proportion of in-person learning to evaluate an association with community-wide daily new SARS-CoV-2 cases. The study period was 12 July 2020-6 October 2020. We included 73 of 92 (79.3%) Indiana counties in the analysis, accounting for 85.7% of school corporations and 90.6% of student enrollment statewide. The primary exposure was the proportion of students returning to in-person instruction. The primary outcome was the daily new SARS-CoV-2 cases per 100 000 residents at the county level. Results: There is a statistically significant relationship between the proportion of students attending K-12 schools in-person and the county level daily cases of SARS-CoV-2 28 days later. For all ages, the coefficient of interest (β) is estimated at 3.36 (95% confidence interval, 1.91 to 4.81; P < .001). This coefficient represents the effect of a change in the proportion of students attending in-person on new daily cases 28 days later. For example, a 10 percentage point increase in K-12 students attending school in-person is associated with a daily increase in SARS-CoV-2 cases in the county equal to 0.336 cases/100 000 residents of all ages. Conclusions: In-person primary and secondary school is associated with a statistically significant but proportionally small increase in the spread of SARS-CoV-2 cases.Item Evaluating Negative Attributions in Persons With Brain Injury: A Comparison of 2 Measures(Wolters Kluwer, 2021-05) Neumann, Dawn; Sander, Angelle M.; Witwer, Noelle; Jang, Jeong Hoon; Bhamidipalli, Surya Sruthi; Hammond, Flora M.; Physical Medicine and Rehabilitation, School of MedicineObjectives: To compare construct and predictive validity, readability, and time-to-administer of 2 negative attribution measures in participants with traumatic brain injury (TBI). Setting: Two TBI rehabilitation hospitals. Participants: Eighty-five adults with complicated mild to severe TBI. Main Measures: Negative attributions (intent, hostility, and blame) and anger responses to hypothetical scenarios were measured with the Epps scenarios and the Ambiguous Intention Hostility Questionnaire (AIHQ). Trait aggression was measured with the Buss-Perry Aggression Questionnaire (BPAQ). Results: Associations between attributions and anger responses (ie, construct validity) within each measure were significant (Epps: r = 0.61-0.74; AIHQ: r = 0.39-0.71); however, associations were stronger for Epps (Ps < .001). Receiver operating characteristics (ROC) revealed attributions from both measures predicted BPAQ scores (area under the ROC curves = 0.6-0.8); predictive validity did not statistically differ between the 2 measures. Both had comparable readability (fifth- to sixth-grade levels), but Epps required longer administration times. Conclusion: Negative attributions affect anger and aggression after TBI, making it important to identify suitable assessments for the TBI population. While psychometric properties of the AIHQ and Epps scenarios should be further explored, this study offers early support for the use of either instrument in persons with TBI. Advantages and disadvantages of the AIHQ and Epps scenarios are highlighted.Item Examination of Social Inferencing Skills in Men and Women After Traumatic Brain Injury(Elsevier, 2022-05) Neumann, Dawn; Mayfield, Ryan; Sander, Angelle M.; Jang, Jeong Hoon; Bhamidipalli, Surya Sruthi; Hammond, Flora M.; Physical Medicine and Rehabilitation, School of MedicineObjective To examine sex differences in social inferencing deficits after traumatic brain injury (TBI) and to examine the odds of men and women being impaired while controlling for potential confounders. Design Cross-sectional survey. Setting Two TBI rehabilitation hospitals. Participants One hundred five participants with TBI (60 men, 45 women) and 105 controls without TBI (57 men, 48 women) (N=210). Interventions Not applicable. Main Outcome Measures The Awareness of Social Inference Test (TASIT), which includes (1) Emotion Evaluation Test (EET), (2) Social Inference-Minimal (SI-M) test, and (3) Social Inference-Enriched (SI-E) test. Results Within the control sample, men and women performed similarly on all 3 TASIT subtests. Within the group with TBI, men had significantly lower scores than women on EET (P=.03), SI-M (P=.01), and SI-E (P=.04). Using impairment cutoffs derived from the sample without TBI, we found significantly more men with TBI (30%) were impaired on the EET than women (16.7%); impairment was similar between men and women on SI-M and SI-E. When adjusting for executive functioning and education, the odds of being impaired on the EET did not significantly differ for men and women (odds ratio, 0.47; 95% CI, 0.16-1.40; P=.18). Conclusions Although more men with TBI have emotion perception deficits than women, the difference appears to be driven by education and executive functioning. Research is needed in larger samples with more definitive norms to better understand social inferencing impairments in men and women with TBI as well as translation to interpersonal behaviors.Item Hydroxyurea reduces infections in children with sickle cell anemia in Uganda(American Society of Hematology, 2024) Namazzi, Ruth; Bond, Caitlin; Conroy, Andrea L.; Datta, Dibyadyuti; Tagoola, Abner; Goings, Michael J.; Jang, Jeong Hoon; Ware, Russell E.; Opoka, Robert; John, Chandy C.; Pediatrics, School of MedicineAfter starting hydroxyurea treatment, Ugandan children with sickle cell anemia had 60% fewer severe or invasive infections, including malaria, bacteremia, respiratory tract infections, and gastroenteritis, than before starting hydroxyurea treatment (incidence rate ratio, 0.40 [ 95% confidence interval, 0.29-0.54 ]; P < .001 ).Item Principal component analysis of hybrid functional and vector data(Wiley, 2021) Jang, Jeong Hoon; Biostatistics and Health Data Science, School of MedicineWe propose a practical principal component analysis (PCA) framework that provides a nonparametric means of simultaneously reducing the dimensions of and modeling functional and vector (multivariate) data. We first introduce a Hilbert space that combines functional and vector objects as a single hybrid object. The framework, termed a PCA of hybrid functional and vector data (HFV-PCA), is then based on the eigen-decomposition of a covariance operator that captures simultaneous variations of functional and vector data in the new space. This approach leads to interpretable principal components that have the same structure as each observation and a single set of scores that serves well as a low-dimensional proxy for hybrid functional and vector data. To support practical application of HFV-PCA, the explicit relationship between the hybrid PC decomposition and the functional and vector PC decompositions is established, leading to a simple and robust estimation scheme where components of HFV-PCA are calculated using the components estimated from the existing functional and classical PCA methods. This estimation strategy allows flexible incorporation of sparse and irregular functional data as well as multivariate functional data. We derive the consistency results and asymptotic convergence rates for the proposed estimators. We demonstrate the efficacy of the method through simulations and analysis of renal imaging data.Item Stepwise Regression and Latent Profile Analyses of Locomotor Outcomes Poststroke(American Heart Association, 2020-10) Hornby, T. George; Henderson, Christopher E.; Holleran, Carey L.; Lovell, Linda; Roth, Elliot J.; Jang, Jeong Hoon; Physical Medicine and Rehabilitation, School of MedicineBackground and purpose: Previous data suggest patient demographics and clinical presentation are primary predictors of motor recovery poststroke, with minimal contributions of physical interventions. Other studies indicate consistent associations between the amount and intensity of stepping practice with locomotor outcomes. The goal of this study was to determine the relative contributions of these combined variables to locomotor outcomes poststroke across a range of patient demographics and baseline function. Methods: Data were pooled from 3 separate trials evaluating the efficacy of high-intensity training, low-intensity training, and conventional interventions. Demographics, clinical characteristics, and training activities from 144 participants >1-month poststroke were included in stepwise regression analyses to determine their relative contributions to locomotor outcomes. Subsequent latent profile analyses evaluated differences in classes of participants based on their responses to interventions. Results: Stepwise regressions indicate primary contributions of stepping activity on locomotor outcomes, with additional influences of age, duration poststroke, and baseline function. Latent profile analyses revealed 2 main classes of outcomes, with the largest gains in those who received high-intensity training and achieved the greatest amounts of stepping practice. Regression and latent profile analyses of only high-intensity training participants indicated age, baseline function, and training activities were primary determinants of locomotor gains. Participants with the smallest gains were older (≈60 years), presented with slower gait speeds (<0.40 m/s), and performed 600 to 1000 less steps/session. Conclusions: Regression and cluster analyses reveal primary contributions of training interventions on mobility outcomes in patients >1-month poststroke. Age, duration poststroke, and baseline impairments were secondary predictors.