Unlocking Direct Lithium Extraction in Harsh Conditions through Thiol-Functionalized Metal-Organic Framework Subnanofluidic Membranes.

Metal-organic framework (MOF) membranes with high ion selectivity are highly desirable for direct lithium-ion (Li+) separation from industrial brines. However, very few MOF membranes can efficiently separate Li+ from brines of high Mg2+/Li+ concentration ratios and keep stable in ultrahigh Mg2+-concentrated brines. This work reports a type of MOF-channel membranes (MOFCMs) by growing UiO-66-(SH)2 into the nanochannels of polymer substrates to improve the efficiency of MOF membranes for challenging Li+ extraction. The resulting membranes demonstrate excellent monovalent metal ion selectivity over divalent metal ions, with Li+/Mg2+ selectivity up to 103 since Mg2+ should overcome a higher energy barrier than Li+ when transported through the MOF pores, as confirmed by molecular dynamics simulations. Under dual-ion diffusion, as the Mg2+/Li+ mole ratio of the feed solution increases from 0.2 to 30, the membrane Li+/Mg2+ selectivity decreases from 1516 to 19, corresponding to the purity of lithium products between 99.9 and 95.0%. Further research on multi-ion diffusion that involves Mg2+ and three monovalent metal ions (K+, Na+, and Li+, referred to as M+) in the feed solutions shows a significant improvement in Li+/Mg2+ separation efficiency. The Li+/Mg2+ selectivity can go up to 1114 when the Mg2+/M+ molar concentration ratio is 1:1, and it remains at 19 when the ratio is 30:1. The membrane selectivity is also stable for 30 days in a highly concentrated solution with a high Mg2+/Li+ concentration ratio. These results indicate the feasibility of the MOFCMs for direct lithium extraction from brines with Mg2+ concentrations up to 3.5 M. This study provides an alternative strategy for designing efficient MOF membranes in extracting valuable minerals in the future.

Combining the Tunnel Orbicularis Oculi Muscle Flap Technique With Skin Grafting for Enhanced Adhesion in Burn-Induced Ectropion Repair.

PURPOSE: Scar contracture of the eyelid following facial burns often has adverse consequences. Total cicatricial contracture often makes adjustment flap translation challenging to implement. Previously used upper and lower eyelid adhesion methods are ineffective for patients with severe cicatricial contracture, and ectropion can easily recur. This study aimed to retrospectively examine upper and lower eyelid adhesions using an orbicularis oculi muscle flap and verify its stability. METHODS: In patients with ectropion caused by severe scar contracture following head and face burns, we employed a tunnel orbicularis oculi muscle flap technique, which involved creating a tunnel between the skin and the tarsal plate of the eyelid, mobilizing the orbicularis oculi muscle, and rotating it into this tunnel to provide stable adhesion of the upper and lower eyelids. Full-thickness skin grafting was then performed. The eyelids were examined postoperatively to determine whether reoperation was necessary and to monitor for any potential complications. RESULTS: This study included 26 patients and 46 eyes. No accidental disconnection occurred after eyelid adhesion, which lasted for an average of 21.87 ± 10.08 months before the eyelid adhesion was cut open. No complications or adverse reactions occurred, and the adhesions did not break unexpectedly. CONCLUSIONS: Repairing eyelid ectropion with the tunnel orbicularis oculi muscle flap is a simple procedure that immediately creates tension against upper and lower eyelid contractures, providing long-term stable adhesion. This method avoids structural disorders, such as eyelid margin scarring, minimally influences surrounding tissues, and has few postoperative complications. It holds great value for repairing eyelid tissue defects and warrants further study.

LATTE: Label-efficient incident phenotyping from longitudinal electronic health records.

Electronic health record (EHR) data are increasingly used to support real-world evidence studies but are limited by the lack of precise timings of clinical events. Here, we propose a label-efficient incident phenotyping (LATTE) algorithm to accurately annotate the timing of clinical events from longitudinal EHR data. By leveraging the pre-trained semantic embeddings, LATTE selects predictive features and compresses their information into longitudinal visit embeddings through visit attention learning. LATTE models the sequential dependency between the target event and visit embeddings to derive the timings. To improve label efficiency, LATTE constructs longitudinal silver-standard labels from unlabeled patients to perform semi-supervised training. LATTE is evaluated on the onset of type 2 diabetes, heart failure, and relapses of multiple sclerosis. LATTE consistently achieves substantial improvements over benchmark methods while providing high prediction interpretability. The event timings are shown to help discover risk factors of heart failure among patients with rheumatoid arthritis.

A Novel Voltage Controlled Decoupling Method for Transmit Coils in MRI.

Array coils are ubiquitous in MRI and are becoming more widely used in MR spectroscopy. Conventional PIN diode decoupling circuits require significant currents to forward bias the diodes. The approach proposed here does not require significant current and thus reduces concerns for contaminating the B0 homogeneity with the detune current. Additionally, the proposed approach will facilitate the construction of array coils for MRI due to its simplicity.Clinical Relevance- Decoupling is critical for constructing RF coil arrays and enables rapid MR imaging.

A Four-Channel Broadband MRI Receive Array Coil.

Low-impedance preamplifier decoupling is commonly used in RF coil array construction to minimize coupling between elements through mutual impedance. The trap circuit is an essential component in preamp decoupling techniques, but becomes a limiting factor in constructing multi-tuned, multi-nuclear coil arrays. In principle, it is possible to double-tune or multi-tune the trap circuits, but will add complexity and loss. We present a broadband decoupling approach using high impedance preamplifiers. A dual-tuned prototype four-channel array using this approach which targets 2H and 23 Na at 4.7T, has been previously constructed, evaluated and reported. Without any retuning of the array, the same setup is tested at the 23Na and 31P frequencies for 3T. Initial bench measurements and Chemical Shift Imaging (CSI) results are acquired and presented in this study.Clinical Relevance- This study could reduce the complexity of multi-nuclear array coil design.

Low-cost MRI System for Teaching.

Magnetic resonance imaging instrumentation is taught at Texas A&M University through the ECEN 463 course and its graduate level equivalent. This class guides students through several labs where they design their own desktop MRI system using various hardware components and LabVIEW. Because the system uses professional grade equipment, the cost of each lab station is high. As a result, there are only four lab stations available, which limits the class to 32 students. The equipment also contains parts that have become obsolete, inhibiting the ability to maintain the system long term. This project focuses on using easily accessible and more affordable equipment for the MRI system. It can also potentially provide opportunities for remote learning, where students could work on assignments off-campus. Other projects have aimed to design low-cost MRI systems with an emphasis on clinical applications or which require advanced FPGA programming skills or pre-programmed modules. This project will develop the MRI instrumentation with updated off-the-shelf components. The current equipment will be replaced with two Analog Discovery 2 devices, which are low-cost teaching tools. It will also feature inexpensive transmit and receive chains, off-the-shelf gradient amplifiers suitable for teaching, gradient coils for signal localization, and a lighter-weight Halbach magnet. In this stage of the project, projections and images have been captured using a 0.06T permanent magnet. In addition to validating successful system operation, each lab of the course will be integrated with current materials to comply with the new equipment. Hardware and software resources will also be prepared and scaled to meet classroom needs and ensure a smooth transition. The goal of the project is to use the new system starting in the fall 2023 semester.Clinical Relevance- This project shows that low-cost equipment can be implemented into a working MRI system. The intent for this project may be educationally focused, but it shows that extremely light and low-cost systems can be created. It may be reconstructed to have a deployable system that could be used in the field.

Stimuli-Responsive Ion Adsorbents for Sustainable Separation Applications.

Stimuli-responsive ion absorbents (SRIAs) with reversible ion adsorption and desorption properties have recently attracted immense attention due to their outstanding functionalities for sustainable separation applications. Over the past decade, a series of SRIAs that respond to single or multiple external stimuli (e.g., pH, gas, temperature, light, magnetic, and voltage) have been reported to achieve excellent ion adsorption capacity and selectivity while simultaneously allowing for their reusability. In contrast to traditional adsorbents that are mainly regenerated through chemical additives, SRIAs allow for reduced chemical and even chemical-free regeneration capacities, thereby enabling environmentally friendly and energy-efficient separation technologies. In this review, we systematically summarize the materials and strategies reported to date for synthesizing single-, dual-, and multiresponsive ion adsorbents. Following a discourse on the fundamental mechanisms that govern their adsorption and desorption under various external stimuli, we provide a concise discussion of the regeneration capacity and application of these responsive ion adsorbents for sustainable water desalination, toxic ion removal, and valuable ion extract and recovery. Finally, we discuss the challenges in developing and deploying these promising multifunctional responsive ion adsorbents together with strategies to overcome these limitations and provide prospects for their future.

Safety Event Outcomes and Glycemic Control with a Hybrid Closed-Loop System Used by Chinese Adolescents and Adults with Type 1 Diabetes Mellitus.

Background: While evidence supports glycemic control benefits for individuals with type 1 diabetes mellitus (T1DM) using hybrid closed-loop (HCL) systems, HCL automated insulin delivery therapy in China has not been assessed. This study evaluated safety events and effectiveness during HCL system use by Chinese adolescents and adults with T1DM. Methods: Sixty-two participants (n = 12 adolescents with a mean ± standard deviation [SD] of 15.5 ± 1.1 years and n = 50 adults [mean ± SD of 37.6 ± 11.1 years]) with T1DM and baseline A1C of 7.1% ± 1.0% underwent a run-in period (∼2 weeks) using open-loop Manual Mode (sensor-augmented pump) insulin delivery with the MiniMed™ 770G system with the Guardian™ Sensor (3) glucose sensor, followed by a study period (4 weeks) with HCL Auto Mode enabled. Analyses compared continuous glucose monitoring data and insulin delivered during the run-in versus study period (Wilcoxon signed-rank test or t-test). Safety events included rates of severe hypoglycemia and diabetic ketoacidosis (DKA). Results: Compared to baseline run-in, overall Auto Mode use increased time in range (TIR, 70-180 mg/dL) from 75.3% to 80.9% (P < 0.001) and reduced time below range (TBR, <70 mg/dL) from 4.7% to 2.2% (P < 0.001). Subgroup analysis demonstrated that participants (n = 29) with baseline A1C <7.0% had TBR that reduced from 5.6% to 2.0%, while participants (n = 21) with baseline A1C ≥7.5% had time above range (TAR, >180 mg/dL) that reduced from 31.6% to 20.8%. Auto Mode use also increased the percentage achieving combined recommendations for time at sensor glucose ranges (i.e., TIR of >70%, TBR of <4% and TAR of <25%) from 24.2% at baseline to 77.4% at study end. Total daily insulin dose reduced from 42.8 ± 19.8 to 40.7 ± 18.9 U (P = 0.013). There were no severe hypoglycemic, DKA, or serious adverse events. Conclusions: Chinese adolescents and adults, some of whom met target A1C at baseline, safely achieved significantly improved glycemia with 1 month of MiniMed 770G system use when compared to open-loop insulin delivery. ClinicalTrials.gov ID: NCT04663295.

Frequency of the 'Asia type' DEL with weak D phenotype in chinese.

BACKGROUND: DEL individuals account for 9-30% of serological RhD negative population in east Asia and majority of them carrying the RHD*DEL1 allele are referred to as 'Asia type' DEL individuals. There is a lack of data on the molecular basis for 'Asia type' DELs with weak RhD phenotype. Therefore, the aim of this study is to unveil 'Asia type' DELs by elucidating the genetic background and analyzing the serological results. METHODS: With a microplate typing protocol, RhD characterization was performed in samples from one million blood donors collected at Chengdu blood center during the period from 2019 to 2022. RhD confirmatory test was performed by direct antiglobulin test and indirect antiglobulin test with five anti-D reagents to detect RhD variants. Molecular characterization of samples categorized as RhD variants was studied by using direct genomic DNA sequencing and RHD zygosity analysis, followed by adsorption and elution tests conduced for samples carrying RHD*DEL1 allele to confirm the presence of RhD antigens on the red cells. RESULTS: We reported here 21 RhD variant samples were detected by micro-column gel agglutination assay with IgG anti-D antibodies. Moreover, the agglutination reaction was stronger with IgG anti-D reagents in micro-column gel card than with IgM/IgG blended anti-D antibodies. Each of the 21 samples carried the RHD*DEL1 allele, which indicated that they were 'Asia type' DEL. Of the 21 'Asia type' DEL samples, 9 samples were detected to be RHD+/ RHD + homozygotes, whereas the other 12 samples were RHD+/RHD- hemizygotes. Among the samples phenotyped for RhCE, seven were CCee and four were Ccee. CONCLUSIONS: In this study, DEL samples carrying RHD*DEL1 showed weak RhD phenotype with some anti-D reagents in RhD confirmatory test, which suggest that a serology strategy using several anti-D reagents may be helpful to detect this 'Asia type' DEL. Further studies are needed to elucidate whether the 'Asia type' DELs with weak RhD phenotype have stronger antigenicity and could cause serious transfusion reaction.

Wideband receive-coil array design using high-impedance amplifiers for broadband decoupling.

PURPOSE: Multinuclear MRI/S is of increasing interest. Currently, most multinuclear receive array coils are constructed by nesting multiple single-tuned array coils or using switching elements to control the operating frequency, in which case more than one set of conventional isolation preamplifiers and associated decoupling circuits is required. These conventional configurations rapidly become complicated when greater numbers of channels or nuclei are needed. In this work, a novel coil decoupling mechanism is proposed to enable broadband decoupling for array coils with one set of preamplifiers. METHODS: Instead of using conventional isolation preamplifiers, a high-input impedance preamplifier is proposed to create broadband decoupling of the array elements. A matching network consisting of a single inductor-capacitor-capacitor multi-tuned network and a wire-wound transformer was used to interface the surface coil to the high-impedance preamplifier. To validate the concept, the proposed configuration was compared to the conventional preamplifier decoupling configuration on both bench and scanner. RESULTS: 2 The approach can provide more than 15dB decoupling over a range of 25MHz, covering the Larmor frequencies of 23 Na and 2 H at 4.7T. This multi-tuned prototype obtained 61% and 76% of the imaging SNR at 2 H and 23 Na respectively, 76 and 89% in a higher loading test phantom, when compared to the conventional single-tuned preamplifier decoupling configuration. CONCLUSION: With the multinuclear array operation and decoupling achieved using only one layer of array coil and preamplifiers, this work provides a simple approach of building high element-count arrays to enable accelerated imaging or SNR improvement from multiple nuclei.

Generate Analysis-Ready Data for Real-world Evidence: Tutorial for Harnessing Electronic Health Records With Advanced Informatic Technologies.

Although randomized controlled trials (RCTs) are the gold standard for establishing the efficacy and safety of a medical treatment, real-world evidence (RWE) generated from real-world data has been vital in postapproval monitoring and is being promoted for the regulatory process of experimental therapies. An emerging source of real-world data is electronic health records (EHRs), which contain detailed information on patient care in both structured (eg, diagnosis codes) and unstructured (eg, clinical notes and images) forms. Despite the granularity of the data available in EHRs, the critical variables required to reliably assess the relationship between a treatment and clinical outcome are challenging to extract. To address this fundamental challenge and accelerate the reliable use of EHRs for RWE, we introduce an integrated data curation and modeling pipeline consisting of 4 modules that leverage recent advances in natural language processing, computational phenotyping, and causal modeling techniques with noisy data. Module 1 consists of techniques for data harmonization. We use natural language processing to recognize clinical variables from RCT design documents and map the extracted variables to EHR features with description matching and knowledge networks. Module 2 then develops techniques for cohort construction using advanced phenotyping algorithms to both identify patients with diseases of interest and define the treatment arms. Module 3 introduces methods for variable curation, including a list of existing tools to extract baseline variables from different sources (eg, codified, free text, and medical imaging) and end points of various types (eg, death, binary, temporal, and numerical). Finally, module 4 presents validation and robust modeling methods, and we propose a strategy to create gold-standard labels for EHR variables of interest to validate data curation quality and perform subsequent causal modeling for RWE. In addition to the workflow proposed in our pipeline, we also develop a reporting guideline for RWE that covers the necessary information to facilitate transparent reporting and reproducibility of results. Moreover, our pipeline is highly data driven, enhancing study data with a rich variety of publicly available information and knowledge sources. We also showcase our pipeline and provide guidance on the deployment of relevant tools by revisiting the emulation of the Clinical Outcomes of Surgical Therapy Study Group Trial on laparoscopy-assisted colectomy versus open colectomy in patients with early-stage colon cancer. We also draw on existing literature on EHR emulation of RCTs together with our own studies with the Mass General Brigham EHR.

Bio-Inspired Subnanofluidics: Advanced Fabrication and Functionalization.

Biological ion channels can realize high-speed and high-selective ion transport through the protein filter with the sub-1-nanometer channel. Inspired by biological ion channels, various kinds of artificial subnanopores, subnanochannels, and subnanoslits with improved ion selectivity and permeability are recently developed for efficient separation, energy conversion, and biosensing. This review article discusses the advanced fabrication and functionalization methods for constructing subnanofluidic pores, channels, tubes, and slits, which have shown great potential for various applications. Novel fabrication methods for producing subnanofluidics, including top-down techniques such as electron beam etching, ion irradiation, and electrochemical etching, as well as bottom-up approaches starting from advanced microporous frameworks, microporous polymers, lipid bilayer embedded subnanochannels, and stacked 2D materials are well summarized. Meanwhile, the functionalization methods of subnanochannels are discussed based on the introduction of functional groups, which are classified into direct synthesis, covalent bond modifications, and functional molecule fillings. These methods have enabled the construction of subnanochannels with precise control of structure, size, and functionality. The current progress, challenges, and future directions in the field of subnanofluidic are also discussed.

Autofluorescence imaging permits label-free cell type assignment and reveals the dynamic formation of airway secretory cell associated antigen passages (SAPs).

The specific functional properties of a tissue are distributed amongst its component cell types. The various cells act coherently, as an ensemble, in order to execute a physiologic response. Modern approaches for identifying and dissecting novel physiologic mechanisms would benefit from an ability to identify specific cell types in live tissues that could then be imaged in real time. Current techniques require the use of fluorescent genetic reporters that are not only cumbersome, but which only allow the study of three or four cell types at a time. We report a non-invasive imaging modality that capitalizes on the endogenous autofluorescence signatures of the metabolic cofactors NAD(P)H and FAD. By marrying morphological characteristics with autofluorescence signatures, all seven of the airway epithelial cell types can be distinguished simultaneously in mouse tracheal explants in real time. Furthermore, we find that this methodology for direct cell type-specific identification avoids pitfalls associated with the use of ostensibly cell type-specific markers that are, in fact, altered by clinically relevant physiologic stimuli. Finally, we utilize this methodology to interrogate real-time physiology and identify dynamic secretory cell associated antigen passages (SAPs) that form in response to cholinergic stimulus. The identical process has been well documented in the intestine where the dynamic formation of SAPs and goblet cell associated antigen passages (GAPs) enable luminal antigen sampling. Airway secretory cells with SAPs are frequently juxtaposed to antigen presenting cells, suggesting that airway SAPs, like their intestinal counterparts, not only sample antigen but convey their cargo for immune cell processing.

Imaging several cell types, at the same time, within a living tissue is no small endeavor. To do so, scientists usually have to perform genetic manipulations that make certain proteins in each cell type fluorescent and therefore easy to track. However, these approaches are cumbersome, limited, and often not applicable to intact human tissues. A possible alternative would be to make use of autofluorescence ­ the fact that certain molecules in living cells naturally fluoresce when exposed to a particular wavelength of light. For example, this is the case for NAD(P)H and FAD, two small molecules necessary for life's biochemical processes, and whose intracellular levels and locations vary depending on cell type. In response, Shah, Hou et al. developed a new imaging technique that takes advantage of the unique autofluorescence signatures of NAD(P)H and FAD to distinguish between the seven different types of cells that line the surface of the airways of mice. Using their autofluorescence approach, Shah, Hou et al. were also able to discover a new role for secretory cells, which normally produce fluids, mucus and various proteins necessary for the lungs to work properly. The imaging experiments show that these cells could also sample material from the surface of the airway in a manner similar to how cells in the intestine take up material from the gut and pass their cargo to immune cells that mediate infection control or tolerance. Further studies should uncover more details about this new function of secretory lung cells. Other exciting discoveries may also emerge from researchers adopting the method developed by Shah, Hou et al. to examine a range of organs (both healthy and diseased), and attempting to apply it to human tissues.

Construction of angstrom-scale ion channels with versatile pore configurations and sizes by metal-organic frameworks.

Controllable fabrication of angstrom-size channels has been long desired to mimic biological ion channels for the fundamental study of ion transport. Here we report a strategy for fabricating angstrom-scale ion channels with one-dimensional (1D) to three-dimensional (3D) pore structures by the growth of metal-organic frameworks (MOFs) into nanochannels. The 1D MIL-53 channels of flexible pore sizes around 5.2 × 8.9 Å can transport cations rapidly, with one to two orders of magnitude higher conductivities and mobilities than MOF channels of hybrid pore configurations and sizes, including Al-TCPP with 1D ~8 Å channels connected by 2D ~6 Å interlayers, and 3D UiO-66 channels of ~6 Å windows and 9 - 12 Å cavities. Furthermore, the 3D MOF channels exhibit better ion sieving properties than those of 1D and 2D MOF channels. Theoretical simulations reveal that ion transport through 2D and 3D MOF channels should undergo multiple dehydration-rehydration processes, resulting in higher energy barriers than pure 1D channels. These findings offer a platform for studying ion transport properties at angstrom-scale confinement and provide guidelines for improving the efficiency of ionic separations and nanofluidics.

Exploring the impact of PEGylation on the cell-nanomicelle interactions by AFM-based single-molecule force spectroscopy and force tracing.

PEGylation has been considered the gold standard method for the modification of various drug delivery systems since the last century. However, the impact of PEGylation on the dynamic interaction between drug carriers and cell membranes has not been quantitatively clarified. Herein, the cellular binding and receptor-mediated endocytosis of a model PEGylated polypeptide nanomicelle were systematically investigated at the single-particle level using AFM-based single-molecule force spectroscopy (SMFS) and force tracing. A self-assembled elastin-like polypeptide (ELP) nanomicelle, which is capable of cross-linking, gastrin-releasing peptide (GRP) modification, and PEGylation was prepared. The cross-linked ELP-based nanomicelles exhibited outstanding stability in a broad temperature range of 4-40 °C, which facilitate the drug loading, as well as our cell-nanomicelle study at the single particle level. The unbinding force between the cross-linked ELP-based nanomicelles and the GRP receptor (GRPR)-containing cell (PC-3) membranes was quantitatively measured by AFM-SMFS. It is found that the PEGylated GRP-displaying nanomicelles exhibit the highest unbinding force, indicating the enhanced specific binding effect of PEGylation. Furthermore, the receptor-mediated endocytosis of the cross-linked ELP-based nanomicelles was monitored with the help of force tracing based on AFM-SMFS. Our results show that PEGylation decreases the endocytic force, duration, and engulfment depth of the PEGylated GRP-displaying nanomicelles, but increases their endocytic velocity, which results from the elimination of non-specific interactions during endocytosis. These observations demonstrate the diverse and complex roles of PEGylation on the interaction of polypeptide nanomicelles to cell membranes and may shed light on the rational design of organic polymer-based drug delivery systems aiming for active and passive targeting strategies. STATEMENT OF SIGNIFICANCE: A self-assembled elastin-like polypeptide (ELP) nanomicelle, which can be easily cross-linked, gastrin-releasing peptide (GRP) modified, and PEGylated, is designed. The AFM-SMFS experiment shows that PEGylation can enhance specific binding of the nanomicelles to the receptors on cell membranes. The force tracing experiment indicates that PEGylation decreases the endocytic force as well as engulfment depth of the nanomicelles through the elimination of non-specific interactions. PEGylation can benefit the drug delivery systems aiming at active targeting, while might not be an ideal modification for drug carriers designed for passive targeting, whose cellular uptake mainly depends on non-specific interactions.

Risk prediction with imperfect survival outcome information from electronic health records.

Readily available proxies for the time of disease onset such as the time of the first diagnostic code can lead to substantial risk prediction error if performing analyses based on poor proxies. Due to the lack of detailed documentation and labor intensiveness of manual annotation, it is often only feasible to ascertain for a small subset the current status of the disease by a follow-up time rather than the exact time. In this paper, we aim to develop risk prediction models for the onset time efficiently leveraging both a small number of labels on the current status and a large number of unlabeled observations on imperfect proxies. Under a semiparametric transformation model for onset and a highly flexible measurement error model for proxy onset time, we propose the semisupervised risk prediction method by combining information from proxies and limited labels efficiently. From an initially estimator solely based on the labeled subset, we perform a one-step correction with the full data augmenting against a mean zero rank correlation score derived from the proxies. We establish the consistency and asymptotic normality of the proposed semisupervised estimator and provide a resampling procedure for interval estimation. Simulation studies demonstrate that the proposed estimator performs well in a finite sample. We illustrate the proposed estimator by developing a genetic risk prediction model for obesity using data from Mass General Brigham Healthcare Biobank.

Association Between Inflammation, Incident Heart Failure, and Heart Failure Subtypes in Patients With Rheumatoid Arthritis.

OBJECTIVE: In rheumatoid arthritis (RA), there are limited data on risk factors for the clinical heart failure (HF) subtypes of HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF). This study examined the association between inflammation and incident HF subtypes in RA. Because inflammation changes over time with disease activity, we hypothesized that the effect of inflammation may be stronger at the 5-year follow-up than at the standard 10-year follow-up from general population studies of cardiovascular risk. METHODS: We studied an electronic health record (EHR)-based RA cohort with data pre- and post-RA incidence. We applied a validated approach to identify HF and extract ejection fraction to classify HFrEF and HFpEF. Follow-up started from the RA incidence date (index date) to the earliest occurrence of incident HF, death, last EHR encounter, or 10 years. Baseline inflammation was assessed using erythrocyte sedimentation rate or C-reactive protein values. Covariates included demographic characteristics, established HF risk factors, and RA-related factors. We tested the association between baseline inflammation with incident HF and its subtypes using Cox proportional hazards models. RESULTS: We studied 9,087 patients with RA; 8.2% developed HF during 10 years of follow-up. Elevated inflammation was associated with increased risk for HF at both 5- and 10-year follow-ups (hazard ratio [HR] 1.66, 95% confidence interval [95% CI] 1.12-2.46 and HR 1.46, 95% CI 1.13-1.90, respectively), which is also seen for HFpEF at 5 years (HR 1.72, 95% CI 1.09-2.70) and 10 years (HR 1.45, 95% CI 1.07-1.94). HFrEF was not associated with inflammation for either follow-up time. CONCLUSION: Elevated inflammation early in RA diagnosis was associated with HF; this association was driven by HFpEF and not HFrEF, suggesting a window of opportunity for prevention of HFpEF in RA.

Surrogate Assisted Semi-supervised Inference for High Dimensional Risk Prediction.

Risk modeling with electronic health records (EHR) data is challenging due to no direct observations of the disease outcome and the high-dimensional predictors. In this paper, we develop a surrogate assisted semi-supervised learning approach, leveraging small labeled data with annotated outcomes and extensive unlabeled data of outcome surrogates and high-dimensional predictors. We propose to impute the unobserved outcomes by constructing a sparse imputation model with outcome surrogates and high-dimensional predictors. We further conduct a one-step bias correction to enable interval estimation for the risk prediction. Our inference procedure is valid even if both the imputation and risk prediction models are misspecified. Our novel way of ultilizing unlabelled data enables the high-dimensional statistical inference for the challenging setting with a dense risk prediction model. We present an extensive simulation study to demonstrate the superiority of our approach compared to existing supervised methods. We apply the method to genetic risk prediction of type-2 diabetes mellitus using an EHR biobank cohort.

Semi-supervised approach to event time annotation using longitudinal electronic health records.

Large clinical datasets derived from insurance claims and electronic health record (EHR) systems are valuable sources for precision medicine research. These datasets can be used to develop models for personalized prediction of risk or treatment response. Efficiently deriving prediction models using real world data, however, faces practical and methodological challenges. Precise information on important clinical outcomes such as time to cancer progression are not readily available in these databases. The true clinical event times typically cannot be approximated well based on simple extracts of billing or procedure codes. Whereas, annotating event times manually is time and resource prohibitive. In this paper, we propose a two-step semi-supervised multi-modal automated time annotation (MATA) method leveraging multi-dimensional longitudinal EHR encounter records. In step I, we employ a functional principal component analysis approach to estimate the underlying intensity functions based on observed point processes from the unlabeled patients. In step II, we fit a penalized proportional odds model to the event time outcomes with features derived in step I in the labeled data where the non-parametric baseline function is approximated using B-splines. Under regularity conditions, the resulting estimator of the feature effect vector is shown as root-n consistent. We demonstrate the superiority of our approach relative to existing approaches through simulations and a real data example on annotating lung cancer recurrence in an EHR cohort of lung cancer patients from Veteran Health Administration.