Search for: All records

Abstract ObjectiveThis study aims to develop machine learning models that provide both accurate and equitable predictions of 2-year stroke risk for patients with atrial fibrillation across diverse racial groups. Materials and MethodsOur study utilized structured electronic health records (EHR) data from the All of Us Research Program. Machine learning models (LightGBM) were utilized to capture the relations between stroke risks and the predictors used by the widely recognized CHADS2 and CHA2DS2-VASc scores. We mitigated the racial disparity by creating a representative tuning set, customizing tuning criteria, and setting binary thresholds separately for subgroups. We constructed a hold-out test set that not only supports temporal validation but also includes a larger proportion of Black/African Americans for fairness validation. ResultsCompared to the original CHADS2 and CHA2DS2-VASc scores, significant improvements were achieved by modeling their predictors using machine learning models (Area Under the Receiver Operating Characteristic curve from near 0.70 to above 0.80). Furthermore, applying our disparity mitigation strategies can effectively enhance model fairness compared to the conventional cross-validation approach. DiscussionModeling CHADS2 and CHA2DS2-VASc risk factors with LightGBM and our disparity mitigation strategies achieved decent discriminative performance and excellent fairness performance. In addition, this approach can provide a complete interpretation of each predictor. These highlight its potential utility in clinical practice. ConclusionsOur research presents a practical example of addressing clinical challenges through the All of Us Research Program data. The disparity mitigation framework we proposed is adaptable across various models and data modalities, demonstrating broad potential in clinical informatics. 

Abstract Microbiome data from sequencing experiments contain the relative abundance of a large number of microbial taxa with their evolutionary relationships represented by a phylogenetic tree. The compositional and high-dimensional nature of the microbiome mediator challenges the validity of standard mediation analyses. We propose a phylogeny-based mediation analysis method called PhyloMed to address this challenge. Unlike existing methods that directly identify individual mediating taxa, PhyloMed discovers mediation signals by analyzing subcompositions defined on the phylogenic tree. PhyloMed produces well-calibrated mediation test p -values and yields substantially higher discovery power than existing methods. 

Individualized treatment rules (ITRs) for treatment recommendation is an important topic for precision medicine as not all beneficial treatments work well for all individuals. Interpretability is a desirable property of ITRs, as it helps practitioners make sense of treatment decisions, yet there is a need for ITRs to be flexible to effectively model complex biomedical data for treatment decision making. Many ITR approaches either focus on linear ITRs, which may perform poorly when true optimal ITRs are nonlinear, or blackbox nonlinear ITRs, which may be hard to interpret and can be overly complex. This dilemma indicates a tension between interpretability and accuracy of treatment decisions. Here we propose an additive model-based nonlinear ITR learning method that balances interpretability and flexibility of the ITR. Our approach aims to strike this balance by allowing both linear and nonlinear terms of the covariates in the final ITR. Our approach is parsimonious in that the nonlinear term is included in the final ITR only when it substantially improves the ITR performance. To prevent overfitting, we combine crossfitting and a specialized information criterion for model selection. Through extensive simulations we show that our methods are data-adaptive to the degree of nonlinearity and can favorably balance ITR interpretability and flexibility. We further demonstrate the robust performance of our methods with an application to a cancer drug sensitive study. 

Abstract The design of high‐entropy single‐atom catalysts (HESAC) with 5.2 times higher entropy compared to single‐atom catalysts (SAC) is proposed, by using four different metals (FeCoNiRu‐HESAC) for oxygen reduction reaction (ORR). Fe active sites with intermetallic distances of 6.1 Å exhibit a low ORR overpotential of 0.44 V, which originates from weakening the adsorption of OH intermediates. Based on density functional theory (DFT) findings, the FeCoNiRu‐HESAC with a nitrogen‐doped sample were synthesized. The atomic structures are confirmed with X‐ray photoelectron spectroscopy (XPS), X‐ray absorption (XAS), and scanning transmission electron microscopy (STEM). The predicted high catalytic activity is experimentally verified, finding that FeCoNiRu‐HESAC has overpotentials of 0.41 and 0.37 V with Tafel slopes of 101 and 210 mVdec⁻¹at the current density of 1 mA cm⁻²and the kinetic current densities of 8.2 and 5.3 mA cm⁻², respectively, in acidic and alkaline electrolytes. These results are comparable with Pt/C. The FeCoNiRu‐HESAC is used for Zinc–air battery applications with an open circuit potential of 1.39 V and power density of 0.16 W cm⁻². Therefore, a strategy guided by DFT is provided for the rational design of HESAC which can be replaced with high‐cost Pt catalysts toward ORR and beyond. 

In this paper, we demonstrate how machine learning could be used to quickly assess a student’s multimodal representational thinking. Multimodal representational thinking is the complex construct that encodes how students form conceptual, perceptual, graphical, or mathematical symbols in their mind. The augmented reality (AR) technology is adopted to diversify student’s representations. The AR technology utilized a low-cost, high-resolution thermal camera attached to a smartphone which allows students to explore the unseen world of thermodynamics. Ninth-grade students (N= 314) engaged in a prediction–observation–explanation (POE) inquiry cycle scaffolded to leverage the augmented observation provided by the aforementioned device. The objective is to investigate how machine learning could expedite the automated assessment of multimodal representational thinking of heat energy. Two automated text classification methods were adopted to decode different mental representations students used to explain their haptic perception, thermal imaging, and graph data collected in the lab. Since current automated assessment in science education rarely considers multilabel classification, we resorted to the help of the state-of-the-art deep learning technique—bidirectional encoder representations from transformers (BERT). The BERT model classified open-ended responses into appropriate categories with higher precision than the traditional machine learning method. The satisfactory accuracy of deep learning in assigning multiple labels is revolutionary in processing qualitative data. The complex student construct, such as multimodal representational thinking, is rarely mutually exclusive. The study avails a convenient technique to analyze qualitative data that does not satisfy the mutual-exclusiveness assumption. Implications and future studies are discussed.