Inter hospital external validation of interpretable machine learning based triage score for the emergency department using common data model.

Emergency departments (ED) are complex, triage is a main task in the ED to prioritize patient with limited medical resources who need them most. Machine learning (ML) based ED triage tool, Score for Emergency Risk Prediction (SERP), was previously developed using an interpretable ML framework with single center. We aimed to develop SERP with 3 Korean multicenter cohorts based on common data model (CDM) without data sharing and compare performance with inter-hospital validation design. This retrospective cohort study included all adult emergency visit patients of 3 hospitals in Korea from 2016 to 2017. We adopted CDM for the standardized multicenter research. The outcome of interest was 2-day mortality after the patients' ED visit. We developed each hospital SERP using interpretable ML framework and validated inter-hospital wisely. We accessed the performance of each hospital's score based on some metrics considering data imbalance strategy. The study population for each hospital included 87,670, 83,363 and 54,423 ED visits from 2016 to 2017. The 2-day mortality rate were 0.51%, 0.56% and 0.65%. Validation results showed accurate for inter hospital validation which has at least AUROC of 0.899 (0.858-0.940). We developed multicenter based Interpretable ML model using CDM for 2-day mortality prediction and executed Inter-hospital external validation which showed enough high accuracy.

Reversible Photomodulation of Two-Dimensional Electron Gas in LaAlO3/SrTiO3 Heterostructures.

Long-lived photoinduced conductance changes in LaAlO3/SrTiO3 (LAO/STO) heterostructures enable their use in optoelectronic memory applications. However, it remains challenging to quench the persistent photoconductivity (PPC) instantly and reproducibly, which limits the reversible optoelectronic switching. Herein, we demonstrate a reversible photomodulation of two-dimensional electron gas (2DEG) in LAO/STO heterostructures with high reproducibility. By irradiating UV pulses, the 2DEG at the LAO/STO interface is gradually transformed to the PPC state. Notably, the PPC can be completely removed by water treatment when two key requirements are met: (1) the moderate oxygen deficiency in STO and (2) the minimal band edge fluctuation at the interface. Through our X-ray photoelectron spectroscopy and electrical noise analysis, we reveal that the reproducible change in the conductivity of 2DEG is directly attributed to the surface-driven electron relaxation in the STO. Our results provide a stepping-stone toward developing optically tunable memristive devices based on oxide 2DEG systems.

Polymer-Covered Copper Catalysts Alter the Reaction Pathway of the Electrochemical CO2 Reduction Reaction.

The electrochemical CO2 reduction reaction (CO2RR) has attracted considerable attention recently due to the potential conversion of atmospheric CO2 into useful organic products by utilizing electricity from renewable energy sources. However, the selective formation of desired products only via CO2RR has been elusive due to the presence of a myriad of competing reaction pathways, thus calling for effective strategies controlling the reaction coordinates. The control of binding energies of the reaction intermediate, such as *CO, is pivotal to manipulating reaction pathways, and various attempts have been made to accomplish this goal. Herein, we introduce recent endeavors to increase the catalytic selectivity of Cu-based catalysts by surface modification with polymer coating, which can change the local pH, hydrophilicity/hydrophobicity, reaction concentration, etc. The polymer conjugation also contributed to the enhanced electrocatalytic stability of Cu-based catalysts during the CO2RR. We also point to the remaining challenges and provide perspectives on the further development of Cu-polymer hybrid catalysts for the practical CO2RR.

Network modeling of anxiety and psychological characteristics on suicidal behavior: Cross-sectional study.

BACKGROUND: Suicide is influenced by complex interactions among various psychopathological features. We aimed to examine the relationship between suicide risk and psychological risk factors such as defense mechanisms, personality, and anxiety. METHODS: We established a psychiatric database by utilizing the Observational Medical Outcomes Partnership Common Data Model. We conducted a 1:1 propensity score matching with age, sex, and depression severity, and identified a sample (n = 258) with two groups: those with suicidal behavior and those with non-suicidal behavior. Using principal component analysis, we extracted nine psychological scales and applied network analysis to compare relationships among psychological factors between the two groups. RESULTS: Patients with non-suicidal behaviors showed associations between trait anxiety and defense mechanisms, while those with suicidal behaviors did not. For patients with suicidal ideation there was an association between somatization and trait anxiety. Patients with suicide attempts showed associations between paranoia and dissociation connected to trait anxiety. LIMITATIONS: Longitudinal research is required to fully observe transitions from suicidal ideation to attempts and recurrent suicidal events. In addition, these networks may not generalize suicidal behaviors because the group participants are not homogeneous. Lastly, data from self-report questionnaires limits the reliability of responses. CONCLUSIONS: We presented important new insights on suicidal behavior by estimating psychological networks. Patients with non-suicidal behavior may exhibit discrete relationships between defense mechanisms and anxiety, compared to those with suicidal behavior. Patients with suicidal ideation and suicide attempts may show distinct associations between anxiety and psychopathological features.

Attention-Based 3D Human Pose Sequence Refinement Network.

Three-dimensional human mesh reconstruction from a single video has made much progress in recent years due to the advances in deep learning. However, previous methods still often reconstruct temporally noisy pose and mesh sequences given in-the-wild video data. To address this problem, we propose a human pose refinement network (HPR-Net) based on a non-local attention mechanism. The pipeline of the proposed framework consists of a weight-regression module, a weighted-averaging module, and a skinned multi-person linear (SMPL) module. First, the weight-regression module creates pose affinity weights from a 3D human pose sequence represented in a unit quaternion form. Next, the weighted-averaging module generates a refined 3D pose sequence by performing temporal weighted averaging using the generated affinity weights. Finally, the refined pose sequence is converted into a human mesh sequence using the SMPL module. HPR-Net is a simple but effective post-processing network that can substantially improve the accuracy and temporal smoothness of 3D human mesh sequences obtained from an input video by existing human mesh reconstruction methods. Our experiments show that the noisy results of the existing methods are consistently improved using the proposed method on various real datasets. Notably, our proposed method reduces the pose and acceleration errors of VIBE, the existing state-of-the-art human mesh reconstruction method, by 1.4% and 66.5%, respectively, on the 3DPW dataset.

Various skin manifestations related to inflammatory bowel disease: A nationwide cross-sectional study on the Korean population.

Inflammatory bowel disease (IBD) presents with various extraintestinal manifestations. As part of them, various skin diseases are suggested to be related to IBD. We aimed to identify the epidemiology and risk of developing skin manifestations in patients with IBD. We used Korean insurance claims data and selected patients with IBD and age/sex-matched non-IBD subjects between 2013 and 2017 using the diagnosis code and prescription records of IBD-specific medications. The prevalence and risk of concurrent skin diseases were estimated. We identified 64 837 patients with IBD. Reactive skin eruptions including pyoderma gangrenosum and erythema nodosum were associated with IBD with highest odds ratios among three categories of reactive, inflammatory, and autoimmune skin diseases. Inflammatory skin diseases including rosacea, psoriasis, atopic dermatitis, hidradenitis suppurativa, and acne conglobata were significantly associated with IBD, but the association was less marked compared to reactive skin eruptions. The patients with IBD also had a higher risk of autoimmune skin diseases including vitiligo and alopecia areata than non-IBD subjects. We determined that IBD was related to various skin diseases including reactive, inflammatory, and autoimmune skin diseases. Considering these relationships can allow better management of patients with IBD and comorbid skin diseases.

A new treatment protocol of microfocused ultrasound for lower eyelid fat bulging.

BACKGROUND: Microfocused ultrasound (MFU) causes tissue tightening by producing thermal injury zones and is used to treat various age-related changes including lower eyelid fat bulging. OBJECTIVE: To investigate the efficacy of a new treatment protocol of MFU for lower eyelid fat bulging. METHODS AND MATERIALS: We reviewed the medical records of all patients who began MFU for lower eyelid fat bulging from March 2017 to September 2018. MFU was performed in two steps to tighten the lower eyelid dermis and orbital septum. Data on age, sex, bulging severity, and the number of treatment sessions were obtained. Associations of these variables with treatment response were determined through an ordinal logistic regression analysis. RESULTS: Among 191 enrolled patients, 119 (62.3%) and 47 (24.6%) achieved fair and good responses, respectively. In the multivariable analysis, multiple treatment sessions (odds ratio (OR) 6.618; 95% confidence interval (CI) 3.242-13.513; p < .001), moderate bulging (OR 4.328; 95% CI 1.755-10.671; p = .001), and severe bulging (OR 7.570; 95% CI 2.537-22.585; p < .001) were associated with greater treatment response. There were no serious adverse events. CONCLUSIONS: The new treatment protocol of MFU is an effective and safe strategy for lower eyelid fat bulging.

Complete and transparent reporting of primary end points of randomized trials in the dermatology literature: A comparison of registered and published primary end points.

BACKGROUND: Appropriate primary end points in randomized controlled trials (RCTs) improve the quality of the measurement and enable comparison of the findings with those of other trials. OBJECTIVE: To assess the quality of reporting primary end points in RCTs recently published in dermatology journals. METHODS: We identified 134 primary reports of RCTs among original articles in 4 dermatology journals published from January 2016 to December 2018. Details were extracted from articles, supplements, and trial registries. A multivariable logistic regression analysis was conducted to identify factors associated with adequate primary end point reporting. RESULTS: Adequate primary end point reporting was conducted in 76 of 134 RCTs (56.7%). Nine missed the definition of primary end points, and 13 did not define the timing of primary end points in the publications. Among 113 RCTs reporting primary end points explicitly in the articles, 16 showed discrepancies between registration and publication, and 21 were not able to valuate prespecification of primary end points. Multicenter studies and sponsor-initiated trials were significantly associated with adequate reporting quality after adjusting for covariates. LIMITATIONS: Prespecification was evaluated based on a comparison of the article and registry. CONCLUSIONS: The quality of primary end point reporting, particularly in prespecification, has remained unsatisfactory in the recent dermatology literature.

Natural course of new-onset urticaria: Results of a 10-year follow-up, nationwide, population-based study.

BACKGROUND: Previous epidemiologic studies of the natural course of urticaria mainly focused on chronic spontaneous urticaria and were conducted at hospitals. The natural course of new-onset urticaria in the general population is unknown. METHODS: Patients with new-onset urticaria were identified from the National Health Insurance Service-National Sample Cohort data. Patients who had at least one visit for urticaria in 2002 and 2003 were excluded and the study cohort consisted of 1,027,620 subjects with no history of urticaria. We analyzed cumulative incidences of urticaria, chronic urticaria, and chronic urticaria remission using the life table estimation method from 2004 to 2013. Their association with related factors was analyzed using the Cox proportional hazards analysis. RESULTS: From 2004 to 2013, a total of 49,129 patients with new-onset urticaria were identified. The 10-year cumulative incidence rate of urticaria for the general population was 4.9% and that of chronic urticaria among patients with new-onset urticaria was 7.8%. Remission rates of chronic urticaria were 52.6% at 1 year and 88.9% at 5 years. Age, sex, residential area, and autoimmune thyroid disease were significantly associated with urticaria or chronic urticaria, but not with chronic urticaria remission, after adjusting for covariates. Female individuals were more likely to have new-onset urticaria but less likely to develop chronic urticaria compared with male individuals. CONCLUSIONS: During the 10-year follow-up period, only a small proportion of patients with new-onset urticaria developed chronic urticaria. Remission was achieved in the majority of patients with chronic urticaria regardless of demographic characteristics or accompanying thyroid disease.

Misclassification of study designs in the dermatology literature.

BACKGROUND: The appropriate classification of study designs is important for review and assessment of the relevant scientific literature as a basis for decision making; however, little is known about whether study designs have been appropriately reported in the dermatology literature. OBJECTIVE: We aimed to validate the study designs in the dermatology literature and investigate discrepancies between author-reported and actual study designs. METHODS: We reviewed all issues of 3 major dermatology journals from January to December 2016. A total of 295 original articles investigating associations between exposures and health outcomes were included for analysis. We used a validated algorithm to classify the study designs. RESULTS: Among the 295 articles, 174 (59.0%) clearly mentioned the study design in the text. All interventional studies were correctly classified on the basis of study design (n = 42); however, 35 of 132 observational studies (26.5%) showed discrepancies between the author-reported and actual study design. When the author-reported design was a prospective cohort, retrospective cohort, or case-control study (n = 61), approximately half of the studies were misclassified by the authors (n = 30). LIMITATIONS: We analyzed only 3 journals in the dermatology field. CONCLUSIONS: Our findings revealed substantial discrepancies between author-reported and actual study designs in the dermatologic literature, particularly among observational studies.

Constructing an Open-Access Bio-Signal Repository from Intensive Care Units.

Bio-signals can be crucial evidence in detecting urgent clinical events. However, until now, access to this data was limited. We aim to construct and provide a new open bio-signal repository with data gathered from more than 40 intensive care unit (ICU) beds. For doing so, we completed the interfacing system with the patient monitors at the target beds and plan to expand this data set to more than 100 ICU beds. Once completed, we plan to publicly open the data to catalyze interesting clinical-event detection research.