Model-based estimation of individual-level social determinants of health and its applications in All of Us.

OBJECTIVES: We introduce a widely applicable model-based approach for estimating individual-level Social Determinants of Health (SDoH) and evaluate its effectiveness using the All of Us Research Program. MATERIALS AND METHODS: Our approach utilizes aggregated SDoH datasets to estimate individual-level SDoH, demonstrated with examples of no high school diploma (NOHSDP) and no health insurance (UNINSUR) variables. Models are estimated using American Community Survey data and applied to derive individual-level estimates for All of Us participants. We assess concordance between model-based SDoH estimates and self-reported SDoHs in All of Us and examine associations with undiagnosed hypertension and diabetes. RESULTS: Compared to self-reported SDoHs, the area under the curve for NOHSDP is 0.727 (95% CI, 0.724-0.730) and for UNINSUR is 0.730 (95% CI, 0.727-0.733) among the 329 074 All of Us participants, both significantly higher than aggregated SDoHs. The association between model-based NOHSDP and undiagnosed hypertension is concordant with those estimated using self-reported NOHSDP, with a correlation coefficient of 0.649. Similarly, the association between model-based NOHSDP and undiagnosed diabetes is concordant with those estimated using self-reported NOHSDP, with a correlation coefficient of 0.900. DISCUSSION AND CONCLUSION: The model-based SDoH estimation method offers a scalable and easily standardized approach for estimating individual-level SDoHs. Using the All of Us dataset, we demonstrate reasonable concordance between model-based SDoH estimates and self-reported SDoHs, along with consistent associations with health outcomes. Our findings also underscore the critical role of geographic contexts in SDoH estimation and in evaluating the association between SDoHs and health outcomes.

Quantitative multiplex immunohistochemistry reveals inter-patient lymphovascular and immune heterogeneity in primary cutaneous melanoma.

Introduction: Quantitative, multiplexed imaging is revealing complex spatial relationships between phenotypically diverse tumor infiltrating leukocyte populations and their prognostic implications. The underlying mechanisms and tissue structures that determine leukocyte distribution within and around tumor nests, however, remain poorly understood. While presumed players in metastatic dissemination, new preclinical data demonstrates that blood and lymphatic vessels (lymphovasculature) also dictate leukocyte trafficking within tumor microenvironments and thereby impact anti-tumor immunity. Here we interrogate these relationships in primary human cutaneous melanoma. Methods: We established a quantitative, multiplexed imaging platform to simultaneously detect immune infiltrates and tumor-associated vessels in formalin-fixed paraffin embedded patient samples. We performed a discovery, retrospective analysis of 28 treatment-naïve, primary cutaneous melanomas. Results: Here we find that the lymphvasculature and immune infiltrate is heterogenous across patients in treatment naïve, primary melanoma. We categorized five lymphovascular subtypes that differ by functionality and morphology and mapped their localization in and around primary tumors. Interestingly, the localization of specific vessel subtypes, but not overall vessel density, significantly associated with the presence of lymphoid aggregates, regional progression, and intratumoral T cell infiltrates. Discussion: We describe a quantitative platform to enable simultaneous lymphovascular and immune infiltrate analysis and map their spatial relationships in primary melanoma. Our data indicate that tumor-associated vessels exist in different states and that their localization may determine potential for metastasis or immune infiltration. This platform will support future efforts to map tumor-associated lymphovascular evolution across stage, assess its prognostic value, and stratify patients for adjuvant therapy.

Detection of molecular oxygen using nanosecond-laser-induced plasma.

Molecular oxygen (O2) concentration is measured by employing nanosecond laser-induced plasmas (ns-LIP) over a broad temperature spectrum ranging from 300 K to 1000 K, in the presence of an additional oxygen-containing molecule, CO2. Typically, emission spectra emanating from ns-LIP are devoid of molecular information, as the ns-LIP causes the dissociation of molecular species within the plasma. However, atomic oxygen absorption lines that momentarily appear at 777 nm in the broadband emission from the early-stage plasma are determined to be highly sensitive to the O2 mole fraction but negligibly affected by the CO2 mole fraction. The atomic O absorbing the plasma emission originates from the O2 adjacent to the plasma: robust UV radiation from the early-stage plasma selectively dissociates adjacent O2, exhibiting a relatively low photodissociation threshold, thus generating the specific meta-stable oxygen capable of absorbing photons at 777 nm. A theoretical model is introduced, explicating the formation of the meta-stable O atom from adjacent O2. To sustain the UV radiation from the plasma under high-temperature and low-density ambient conditions, a preceding breakdown is triggered by a split laser pulse (532 nm). This breakdown acts as a precursor, seeding electrons to intensify the inverse-Bremsstrahlung photon absorption of the subsequent laser pulse (1064 nm). Techniques such as proper orthogonal decomposition (POD) and support vector regression (SVR) are employed to precisely evaluate the O2 mole fraction (<1% uncertainty), by analyzing the short-lived (<10 ns) O2-indicator depicted in the early-stage plasma.

Gas composition measurements in randomly distributed and fast moving gas bubbles in two-phase fluids.

Gas composition in randomly distributed and fast-moving bubbles was optically measured aided by laser-induced breakdown spectroscopy (LIBS). Laser pulses were focused at a point in a stream of bubbles to induce plasmas for the LIBS measurements. The distance between the laser focal point and liquid-gas interface, or 'depth,' plays a major role in determining the plasma emission spectrum in two-phase fluids. However, the 'depth' effect has not been investigated in previous studies. Therefore, we evaluated the 'depth' effect in a calibration experiment near a still and flat liquid-gas interface using proper orthogonal decomposition, and a support vector regression model was trained to exclude the influence of the interfacing liquid and extract gas composition information from the spectra. The gaseous molecular oxygen mole fraction in the bubbles was accurately measured under realistic two-phase fluid conditions.

A joint fairness model with applications to risk predictions for underrepresented populations.

In data collection for predictive modeling, underrepresentation of certain groups, based on gender, race/ethnicity, or age, may yield less accurate predictions for these groups. Recently, this issue of fairness in predictions has attracted significant attention, as data-driven models are increasingly utilized to perform crucial decision-making tasks. Existing methods to achieve fairness in the machine learning literature typically build a single prediction model in a manner that encourages fair prediction performance for all groups. These approaches have two major limitations: (i) fairness is often achieved by compromising accuracy for some groups; (ii) the underlying relationship between dependent and independent variables may not be the same across groups. We propose a joint fairness model (JFM) approach for logistic regression models for binary outcomes that estimates group-specific classifiers using a joint modeling objective function that incorporates fairness criteria for prediction. We introduce an accelerated smoothing proximal gradient algorithm to solve the convex objective function, and present the key asymptotic properties of the JFM estimates. Through simulations, we demonstrate the efficacy of the JFM in achieving good prediction performance and across-group parity, in comparison with the single fairness model, group-separate model, and group-ignorant model, especially when the minority group's sample size is small. Finally, we demonstrate the utility of the JFM method in a real-world example to obtain fair risk predictions for underrepresented older patients diagnosed with coronavirus disease 2019 (COVID-19).

Fair Survival Time Prediction via Mutual Information Minimization.

Survival analysis is a general framework for predicting the time until a specific event occurs, often in the presence of censoring. Although this framework is widely used in practice, few studies to date have considered fairness for time-to-event outcomes, despite recent significant advances in the algorithmic fairness literature more broadly. In this paper, we propose a framework to achieve demographic parity in survival analysis models by minimizing the mutual information between predicted time-to-event and sensitive attributes. We show that our approach effectively minimizes mutual information to encourage statistical independence of time-to-event predictions and sensitive attributes. Furthermore, we propose four types of disparity assessment metrics based on common survival analysis metrics. Through experiments on multiple benchmark datasets, we demonstrate that by minimizing the dependence between the prediction and the sensitive attributes, our method can systematically improve the fairness of survival predictions and is robust to censoring.

Nanosecond laser pulse modulation using seed electrons from cascade ionization induced by inverse-Bremsstrahlung photon absorption.

Nanosecond (ns) laser pulses are modulated by seeding electrons on the laser beam path. The seed-electrons are from auxiliary ns-laser-induced breakdown (ALIB), and the ALIB is induced by a focused 1064-nm pulse, which is split after the frequency-doubling that generates the 532-nm pulse; therefore, the 532-nm and 1064-nm pulses are synchronized. The slowly converging (focal length = 500 mm) 532-nm pulse is re-directed to transmit through the region in where the ALIB-generated electrons reside. The seed-electrons from the ALIB then absorb the 532-nm photons via the inverse-Bremsstrahlung photon absorption (IBPA) process. The number density of the seed-electrons on the 532-nm beam path (ne,ALIB) is controlled by varying 1) the 532-nm pulse arrival time at the ALIB region (ΔPAT) after the 1064-nm pulse triggers the ALIB and 2) the location of the 532-nm beam relative to the core of the ALIB; the electron number density in ALIB is highly non-uniform and evolves in time. Electron-seeded laser-induced breakdown (ESLIB) occurs when ne,ALIB is sufficiently high. The 532-nm beam convergence (controlled by the focusing lens) is adjusted so that the breakdown does not occur without the electron seeding. The ESLIB immediately stops the transmission of the trailing edge of the laser pulse acting as a fast shutter, and ne,ALIB above a threshold can cut the pulse leading edge to modulate the 532-nm laser pulse.

Factors associated with hospital admission and severe outcomes for older patients with COVID-19.

BACKGROUND: Morbidity and death due to coronavirus disease 2019 (COVID-19) experienced by older adults in nursing homes have been well described, but COVID-19's impact on community-living older adults is less studied. Similarly, the previous ambulatory care experience of such patients has rarely been considered in studies of COVID-19 risks and outcomes. METHODS: To investigate the relationship of advanced age (65+), on risk factors associated with COVID-19 outcomes in community-living elders, we identified an electronic health records cohort of older patients aged 65+ with laboratory-confirmed COVID-19 with and without an ambulatory care visit in the past 24 months (n = 47,219) in the New York City (NYC) academic medical institutions and the NYC public hospital system from January 2020 to February 2021. The main outcomes are COVID-19 hospitalization; severe outcomes/Intensive care unit (ICU), intubation, dialysis, stroke, in-hospital death), and in-hospital death. The exposures include demographic characteristics, and those with ambulatory records, comorbidities, frailty, and laboratory results. RESULTS: The 31,770 patients with an ambulatory history had a median age of 74 years; were 47.4% male, 24.3% non-Hispanic white, 23.3% non-Hispanic black, and 18.4% Hispanic. With increasing age, the odds ratios and attributable fractions of sex, race-ethnicity, comorbidities, and biomarkers decreased except for dementia and frailty (Hospital Frailty Risk Score). Patients without ambulatory care histories, compared to those with, had significantly higher adjusted rates of COVID-19 hospitalization and severe outcomes, with strongest effect in the oldest group. CONCLUSIONS: In this cohort of community-dwelling older adults, we provided evidence of age-specific risk factors for COVID-19 hospitalization and severe outcomes. Future research should explore the impact of frailty and dementia in severe COVID-19 outcomes in community-living older adults, and the role of engagement in ambulatory care in mitigating severe disease.

Fair Generalized Linear Models with a Convex Penalty.

Despite recent advances in algorithmic fairness, methodologies for achieving fairness with generalized linear models (GLMs) have yet to be explored in general, despite GLMs being widely used in practice. In this paper we introduce two fairness criteria for GLMs based on equalizing expected outcomes or log-likelihoods. We prove that for GLMs both criteria can be achieved via a convex penalty term based solely on the linear components of the GLM, thus permitting efficient optimization. We also derive theoretical properties for the resulting fair GLM estimator. To empirically demonstrate the efficacy of the proposed fair GLM, we compare it with other wellknown fair prediction methods on an extensive set of benchmark datasets for binary classification and regression. In addition, we demonstrate that the fair GLM can generate fair predictions for a range of response variables, other than binary and continuous outcomes.

Machine learning based quantification of fuel-air equivalence ratio and pressure from laser-induced plasma spectroscopy.

In this study, we demonstrate successful development of a predictive model that detects both the fuel-air equivalence ratio (ϕ) and local pressure prior to plasma formation via machine-learning from the laser-induced plasma spectra; the resulting model enables measurement of a wide range of fuel concentrations and pressures. The process of model acquisition is composed of three steps: (i) normalization of the spectra, (ii) feature extraction and selection, and (iii) training of an artificial neural network (ANN) with feature scores and the corresponding labels. In detail, the spectra were first normalized by the total emission intensity; then principal component analysis (PCA) or independent component analysis (ICA) was carried out for feature extraction and selection. Subsequently, the scores of these principal or independent components as inputs were trained for the ANN with expected ϕ and pressure values for outputs, respectively. The model acquisition was successful, and the model's predictive performance was validated by predicting the ϕ and pressure in the test dataset.

A Joint Fairness Model with Applications to Risk Predictions for Under-represented Populations.

In data collection for predictive modeling, under-representation of certain groups, based on gender, race/ethnicity, or age, may yield less-accurate predictions for these groups. Recently, this issue of fairness in predictions has attracted significant attention, as data-driven models are increasingly utilized to perform crucial decision-making tasks. Existing methods to achieve fairness in the machine learning literature typically build a single prediction model in a manner that encourages fair prediction performance for all groups. These approaches have two major limitations: i) fairness is often achieved by compromising accuracy for some groups; ii) the underlying relationship between dependent and independent variables may not be the same across groups. We propose a Joint Fairness Model (JFM) approach for logistic regression models for binary outcomes that estimates group-specific classifiers using a joint modeling objective function that incorporates fairness criteria for prediction. We introduce an Accelerated Smoothing Proximal Gradient Algorithm to solve the convex objective function, and present the key asymptotic properties of the JFM estimates. Through simulations, we demonstrate the efficacy of the JFM in achieving good prediction performance and across-group parity, in comparison with the single fairness model, group-separate model, and group-ignorant model, especially when the minority group's sample size is small. Finally, we demonstrate the utility of the JFM method in a real-world example to obtain fair risk predictions for under-represented older patients diagnosed with coronavirus disease 2019 (COVID-19).

Dynamic Survival Analysis for EHR Data with Personalized Parametric Distributions.

The widespread availability of high-dimensional electronic healthcare record (EHR) datasets has led to significant interest in using such data to derive clinical insights and make risk predictions. More specifically, techniques from machine learning are being increasingly applied to the problem of dynamic survival analysis, where updated time-to-event risk predictions are learned as a function of the full covariate trajectory from EHR datasets. EHR data presents unique challenges in the context of dynamic survival analysis, involving a variety of decisions about data representation, modeling, interpretability, and clinically meaningful evaluation. In this paper we propose a new approach to dynamic survival analysis which addresses some of these challenges. Our modeling approach is based on learning a global parametric distribution to represent population characteristics and then dynamically locating individuals on the time-axis of this distribution conditioned on their histories. For evaluation we also propose a new version of the dynamic C-Index for clinically meaningful evaluation of dynamic survival models. To validate our approach we conduct dynamic risk prediction on three real-world datasets, involving COVID-19 severe outcomes, cardiovascular disease (CVD) onset, and primary biliary cirrhosis (PBC) time-to-transplant. We find that our proposed modeling approach is competitive with other well-known statistical and machine learning approaches for dynamic risk prediction, while offering potential advantages in terms of interepretability of predictions at the individual level.

Laser pulse width control using inverse-Bremsstrahlung photon absorption.

Nanosecond laser pulses (6 ns FWHM, produced by a Q-switched, frequency-doubled Nd:YAG laser) are chopped using the inverse-Bremsstrahlung (IB) photon absorption process in a cell with variable pressure. The IB process that quickly absorbs the majority of the laser pulse energy is triggered by focusing the pulse in the cell. Prior to the initiation of the IB process, the gaseous medium in the cell is transparent, while it suddenly becomes opaque with the IB process activated; therefore, the pressure cell can be used as a virtual optical shutter. The shutter "closing time" depends strongly on the pressure of the cell and the laser pulse energy and thus can be controlled. Dependence of the "closing time" on these two parameters is experimentally investigated.

Evolution of Cavitation Bubble in Tap Water by Continuous-Wave Laser Focused on a Metallic Surface.

As an example of photon-matter interaction, we experimentally investigate the temporal evolution of a millimeter-sized cavitation bubble, induced by focusing a continuous-wave laser on a metallic plate in tap water. Our major interests are to understand the mechanism of bubble growth/shrinkage for a long time duration up to O(102) seconds and to draw the time-dependency relation of a bubble size, depending on the incident laser power. With the time passed after the laser with different power is focused, it is found that the phase change and/or transport of dissolved gas into the bubble play a dominant role in determining the bubble growth and shrinkage. Thus, we propose two stages in terms of time and three regimes depending on the incident energy, in which the evolutions of cavitation bubble in short and long time durations are distinctively identified. In regime I (lower incident power), the water nearby the focal point undergoes a phase change, resulting in an initial rapid growth of a bubble (first stage), but the convection flow due to locally heated surface causes the bubble to shrink at later times (second stage). As the laser power increases (regime III), more dissolved gas in the surrounding water enters the growing bubble and prevents the water phase from being absorbed into the bubble. Thus, the bubble growth is dominated by the dissolved gas. Between regimes I and III, there is a transitional regime II in which both the phase change of water and the dissolved gas contribute to the bubble evolution. We further our understandings by developing the relations about the time-dependency of bubble size for each stage and regime, which agree well with the measured data. The scaling relations are also validated with different conditions of liquid such as degassed water and NaCl solution. While previous studies have mostly focused on the nano- and/or microsized bubble generation in a very short time (less than 1 s), we think that the present results will extend our knowledge on how to predict and control the size of laser-induced cavitation bubble for longer time duration.