Hybrid feature engineering of medical data via variational autoencoders with triplet loss: a COVID-19 prognosis study.

Medical machine learning frameworks have received much attention in recent years. The recent COVID-19 pandemic was also accompanied by a surge in proposed machine learning algorithms for tasks such as diagnosis and mortality prognosis. Machine learning frameworks can be helpful medical assistants by extracting data patterns that are otherwise hard to detect by humans. Efficient feature engineering and dimensionality reduction are major challenges in most medical machine learning frameworks. Autoencoders are novel unsupervised tools that can perform data-driven dimensionality reduction with minimum prior assumptions. This study, in a novel approach, investigated the predictive power of latent representations obtained from a hybrid autoencoder (HAE) framework combining variational autoencoder (VAE) characteristics with mean squared error (MSE) and triplet loss for forecasting COVID-19 patients with high mortality risk in a retrospective framework. Electronic laboratory and clinical data of 1474 patients were used in the study. Logistic regression with elastic net regularization (EN) and random forest (RF) models were used as final classifiers. Moreover, we also investigated the contribution of utilized features towards latent representations via mutual information analysis. HAE Latent representations model achieved decent performance with an area under ROC curve of 0.921 (±0.027) and 0.910 (±0.036) with EN and RF predictors, respectively, over the hold-out data in comparison with the raw (AUC EN: 0.913 (±0.022); RF: 0.903 (±0.020)) models. The study aims to provide an interpretable feature engineering framework for the medical environment with the potential to integrate imaging data for efficient feature engineering in rapid triage and other clinical predictive models.

Long-term laryngological sequelae and patient-reported outcomes after COVID-19 infection.

PURPOSE: We examine prevalence, characteristics, quality of life (QOL) assessments, and long-term effects of interventions for laryngeal dysfunction after recovery from COVID-19 infection. MATERIALS AND METHODS: 653 patients presenting to Yale's COVID clinic from April 2020 to August 2021 were identified retrospectively. Patients with PCR-positive COVID-19 who underwent evaluation by fellowship-trained laryngologists were included. Patient demographics, comorbidities, intubation/tracheostomy, strobolaryngoscopy, voice metrics, and management data were collected. Patient-reported QOL indices were Dyspnea Index (DI), Cough Severity Index (CSI), Voice Handicap Index-10 (VHI-10), Eating Assessment Tool-10 (EAT-10), and Reflux Symptom Index (RSI). RESULTS: 57 patients met inclusion criteria: 37 (64.9 %) were hospitalized for COVID-19 infection and 24 (42.1 %) required intubation. Mean duration between COVID-19 diagnosis and presentation to laryngology was significantly shorter for patients who were intubated compared to non-intubated (175 ± 98 days versus 256 ± 150 days, respectively, p = 0.025). Dysphonia was diagnosed in 40 (70.2 %) patients, dysphagia in 14 (25.0 %) patients, COVID-related laryngeal hypersensitivity in 13 (22.8 %), and laryngotracheal stenosis (LTS) in 10 (17.5 %) patients. Of the 17 patients who underwent voice therapy, 11 (64.7 %) reported improvement in their symptoms and 2 (11.8 %) patients reported resolution. VHI scores decreased for patients who reported symptom improvement. 7 (70 %) patients with LTS required >1 procedural intervention before symptom improvement. Improvement across QOL indices was seen in patients with LTS. CONCLUSIONS: Laryngeal dysfunction commonly presents and is persistent for months after recovery from COVID-19 in non-hospitalized and non-intubated patients. Voice therapy and procedural interventions have the potential to address post-COVID laryngeal dysfunction.

Malaria diagnostic methods with the elimination goal in view.

Malaria control measures have been in use for years but have not completely curbed the spread of infection. Ultimately, global elimination is the goal. A major playmaker in the various approaches to reaching the goal is the issue of proper diagnosis. Various diagnostic techniques were adopted in different regions and geographical locations over the decades, and these have invariably produced diverse outcomes. In this review, we looked at the various approaches used in malaria diagnostics with a focus on methods favorably used during pre-elimination and elimination phases as well as in endemic regions. Microscopy, rapid diagnostic testing (RDT), loop-mediated isothermal amplification (LAMP), and polymerase chain reaction (PCR) are common methods applied depending on prevailing factors, each with its strengths and limitations. As the drive toward the elimination goal intensifies, the search for ideal, simple, fast, and reliable point-of-care diagnostic tools is needed more than ever before to be used in conjunction with a functional surveillance system supported by the ideal vaccine.

Low-cost, point-of-care biomarker quantification.

Low-cost, point-of-care (POC) devices that allow fast, on-site disease diagnosis could have a major global health impact, particularly if they can provide quantitative measurement of molecules indicative of a diseased state (biomarkers). Accurate quantification of biomarkers in patient samples is already challenging when research-grade, sophisticated equipment is available; it is even more difficult when constrained to simple, cost-effective POC platforms. Here, we summarize the main challenges to accurate, low-cost POC biomarker quantification. We also review recent efforts to develop and implement POC tools beyond qualitative readouts, and we conclude by identifying important future research directions.

Early anti-S antibody levels predict anti-SARS-CoV-2 neutralizing activity over 24 weeks in RA patients after SARS-CoV-2 mRNA vaccination

Objectives: Routine monitoring of vaccine-induced anti-S responses following mRNA based SARS-CoV-2 vaccination is not recommended routinely as uncertainties exist about the critical threshold of antibody levels that correlate to protection and the optimal timepoint for determination. In our study, anti-S antibody were analysed over 24 weeks following a standard two-dose regimen of mRNA based anti-SARS-CoV-2 vaccines and correlated to the development and persistence of neutralizing activity against SARSCoV- 2 in patients with rheumatoid arthritis (RA) on DMARD therapy compared to healthy controls (HC). Method(s): The RECOVER study was a prospective, controlled, monocentric study. Assessments were performed before vaccination, and at three, six, 12 and 24 weeks after the first vaccine dose. Result(s): In RA patients, anti-S responses developed slower and resulted in lower peak titers compared to HC. A potent neutralizing activity (NT50) as assessed by a SARS-CoV-2 pseudoneutralization assay was observed in 60.3 % of all 73 RA patients and in all 21 HC after 12 weeks. A significant correlation between peak anti-S levels two weeks after the second vaccine dose and the development of a persistent neutralizing activity against SARS-CoV-2 was observed at week 12 and week 24. The analysis of IgG, IgA, and IgM isotype responses to RBD, S1, S2, and N proteins revealed a delayed IgG response, while IgA and IgM responses were maintained, suggesting a delayed isotype switch in RA patients. Conclusion(s): Peak anti-S IgG levels two weeks after the second vaccine dose significantly predicted the development and persistence of a potent neutralizing activity against SARS-CoV-2 after 12 and 24 weeks. Our data suggest that the early determination of anti- S levels allows the timely identification of non- or poor-responding patients.

Trends in Residency Applicant Volume in Otolaryngology-Head and Neck Surgery and Peer Specialties.

OBJECTIVE: To evaluate the 2020 to 2021 Otolaryngology residency application cycle in the context of recent trends. STUDY DESIGN: Retrospective data analysis. SETTING: Disruptions caused by the COVID-19 pandemic may significantly alter trends among residency applicants, especially in highly competitive and/or smaller specialties. METHODS: Applicant and residency statistics from Electronic Residency Application Service (ERAS) and National Residency Matching Program (NRMP) were extracted from the 2016 to 2021 and 2011 to 2021, respectively. Trends in Otolaryngology-Head and Neck Surgery (OHNS) were compared to peer specialties (PS) including Dermatology, Neurological Surgery, Orthopedic Surgery, and Integrated Pathway for Plastic and Reconstructive Surgery (PRS). The ratio of the number of applicants per positions (APP) was used to reflect the degree of competition. RESULTS: Between 2011 and 2021, the number of OHNS programs and positions expanded less than those of PS and General Surgery. The increase in the APP ratio was significantly greater for OHNS compared to those Dermatology, Orthopedic Surgery, General Surgery and all PGY1 residency positions for both US MD and all applicants (P < .01 for each). OHNS expansion of US MD (P = .046), but not all applicants (P = .169), outgrew that of Neurosurgery. CONCLUSION: The 2020 to 2021 cycle affected by the COVID-19 pandemic saw a continuation of the recent trend in the expanding OHNS applicant pool. OHNS remains one of the specialties with the highest APP ratio and has observed a significant growth compared to PS since 2018. Understanding and anticipating trends in residency application cycles is critical for designing processes to optimize the best fit between applicants and programs.

Multisite implementation of a workflow-integrated machine learning system to optimize COVID-19 hospital admission decisions.

Demand has outstripped healthcare supply during the coronavirus disease 2019 (COVID-19) pandemic. Emergency departments (EDs) are tasked with distinguishing patients who require hospital resources from those who may be safely discharged to the community. The novelty and high variability of COVID-19 have made these determinations challenging. In this study, we developed, implemented and evaluated an electronic health record (EHR) embedded clinical decision support (CDS) system that leverages machine learning (ML) to estimate short-term risk for clinical deterioration in patients with or under investigation for COVID-19. The system translates model-generated risk for critical care needs within 24 h and inpatient care needs within 72 h into rapidly interpretable COVID-19 Deterioration Risk Levels made viewable within ED clinician workflow. ML models were derived in a retrospective cohort of 21,452 ED patients who visited one of five ED study sites and were prospectively validated in 15,670 ED visits that occurred before (n = 4322) or after (n = 11,348) CDS implementation; model performance and numerous patient-oriented outcomes including in-hospital mortality were measured across study periods. Incidence of critical care needs within 24 h and inpatient care needs within 72 h were 10.7% and 22.5%, respectively and were similar across study periods. ML model performance was excellent under all conditions, with AUC ranging from 0.85 to 0.91 for prediction of critical care needs and 0.80-0.90 for inpatient care needs. Total mortality was unchanged across study periods but was reduced among high-risk patients after CDS implementation.

Machine learning and artificial intelligence: applications in healthcare epidemiology.

Artificial intelligence (AI) refers to the performance of tasks by machines ordinarily associated with human intelligence. Machine learning (ML) is a subtype of AI; it refers to the ability of computers to draw conclusions (ie, learn) from data without being directly programmed. ML builds from traditional statistical methods and has drawn significant interest in healthcare epidemiology due to its potential for improving disease prediction and patient care. This review provides an overview of ML in healthcare epidemiology and practical examples of ML tools used to support healthcare decision making at 4 stages of hospital-based care: triage, diagnosis, treatment, and discharge. Examples include model-building efforts to assist emergency department triage, predicting time before septic shock onset, detecting community-acquired pneumonia, and classifying COVID-19 disposition risk level. Increasing availability and quality of electronic health record (EHR) data as well as computing power provides opportunities for ML to increase patient safety, improve the efficiency of clinical management, and reduce healthcare costs.

Discovery and functional interrogation of SARS-CoV-2 protein-RNA interactions (preprint)

The COVID-19 pandemic is caused by severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2). The betacoronvirus has a positive sense RNA genome which encodes for several RNA binding proteins. Here, we use enhanced crosslinking and immunoprecipitation to investigate SARS-CoV-2 protein interactions with viral and host RNAs in authentic virus-infected cells. SARS-CoV-2 proteins, NSP8, NSP12, and nucleocapsid display distinct preferences to specific regions in the RNA viral genome, providing evidence for their shared and separate roles in replication, transcription, and viral packaging. SARS-CoV-2 proteins expressed in human lung epithelial cells bind to 4773 unique host coding RNAs. Nine SARS-CoV-2 proteins upregulate target gene expression, including NSP12 and ORF9c, whose RNA substrates are associated with pathways in protein N-linked glycosylation ER processing and mitochondrial processes. Furthermore, siRNA knockdown of host genes targeted by viral proteins in human lung organoid cells identify potential antiviral host targets across different SARS-CoV-2 variants. Conversely, NSP9 inhibits host gene expression by blocking mRNA export and dampens cytokine productions, including interleukin-1/{beta}. Our viral protein-RNA interactome provides a catalog of potential therapeutic targets and offers insight into the etiology of COVID-19 as a safeguard against future pandemics.

Abnormal liver chemistries as a predictor of COVID-19 severity and clinical outcomes in hospitalized patients.

BACKGROUND: Abnormal liver chemistries are common findings in patients with Coronavirus Disease 2019 (COVID-19). However, the association of these abnormalities with the severity of COVID-19 and clinical outcomes is poorly understood. AIM: We aimed to assess the prevalence of elevated liver chemistries in hospitalized patients with COVID-19 and compare the serum liver chemistries to predict the severity and in-hospital mortality. METHODS: This retrospective, observational study included 3380 patients with COVID-19 who were hospitalized in the Johns Hopkins Health System (Baltimore, MD, United States). Demographic data, clinical characteristics, laboratory findings, treatment measures, and outcome data were collected. Cox regression modeling was used to explore variables associated with abnormal liver chemistries on admission with disease severity and prognosis. RESULTS: A total of 2698 (70.4%) had abnormal alanine aminotransferase (ALT) at the time of admission. Other more prevalent abnormal liver chemistries were aspartate aminotransferase (AST) (44.4%), alkaline phosphatase (ALP) (16.1%), and total bilirubin (T-Bil) (5.9%). Factors associated with liver injury were older age, Asian ethnicity, other race, being overweight, and obesity. Higher ALT, AST, T-Bil, and ALP levels were more commonly associated with disease severity. Multivariable adjusted Cox regression analysis revealed that abnormal AST and T-Bil were associated with the highest mortality risk than other liver injury indicators during hospitalization. Abnormal AST, T-Bil, and ALP were associated with a need for vasopressor drugs, whereas higher levels of AST, T-Bil, and a decreased albumin levels were associated with mechanical ventilation. CONCLUSION: Abnormal liver chemistries are common at the time of hospital admission in COVID-19 patients and can be closely related to the patient's severity and prognosis. Elevated liver chemistries, specifically ALT, AST, ALP, and T-Bil levels, can be used to stratify risk and predict the need for advanced therapies in these patients.

Cell delivery peptides for small interfering RNAs targeting SARS-CoV-2 new variants through a bioinformatics and deep learning design (preprint)

Nucleic acid technologies with designed delivery systems have surged as one the most promising therapies of the future, due to their contribution in combating SARS-CoV-2 severe disease. Nevertheless, the emergence of new variants of concern still represents a real threat in the years to come. It is here that the use of small interfering RNA sequences to inhibit gene expression and, thus, protein synthesis, may complement the already developed vaccines, with faster design and production. Here, we have designed new sequences targeting COVID-19 variants and other related viral diseases through bioinformatics, while also addressing the limited number of delivery peptides by a deep learning approach. Two sequences databases were produced, from which 62 were able to target the virus mRNA, and ten displayed properties present in delivery peptides, which we compared to the broad use TAT delivery peptide.

Towards accessibility in education through smart speakers. An ontology based approach

As the world changes, so does the future of our students. In this respect, the evolution of the technology comes up with specific environments for educational purpose. Building smart learning environments supported by e-learning platforms is an important area of research in education domain within our days. The evolution of these smart learning environments is justified by some events (Covid19) that force students to learn remotely. The paper proposes a formalisation using ontology for providing an inclusive approach of universities' websites, having as instance a software application component using Alexa smart speaker, that currently remains at a design level, which integrates different services (Amazon Web Services, Microsoft Services) for a proper virtual environment platform, for both students and teachers. It addresses the main concerns of the current educational system and provides a smart solution through the use of Artificial Intelligence based tools. The proposed approach not only achieves unifying data and knowledge-share mechanisms in a remotely mode, but it brings also a good learning experience, increasing the effectiveness and the efficiency of the learning process. © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of KES International.

Heart Rate Reactivity to Acute Psychological Stress Predicts Higher Levels of Posttraumatic Stress Disorder Symptoms During the COVID-19 Pandemic.

OBJECTIVE: Increased autonomic arousal is a proposed risk factor for posttraumatic stress disorder (PTSD). Few studies have prospectively examined the association between physiological responses to acute psychological stress before a traumatic event and later PTSD symptoms. The present prospective study examined whether cardiovascular responses to an acute psychological stress task before the COVID-19 global pandemic predicted PTSD symptoms related to the ongoing pandemic. METHODS: Participants (n = 120) were a subsample of an ongoing research study. Phase 1 consisted of a 10-minute baseline and 4-minute acute psychological stress task with blood pressure and heart rate recorded throughout. Phase 2 was initiated 2 weeks after the COVID-19 pandemic declaration. Participants completed the Impact of Event Scale-Revised (IES-R) with respect to the ongoing pandemic. Hierarchical linear regression analyses were used to examine whether cardiovascular stress reactivity predicted COVID-19 PTSD symptoms. RESULTS: Heart rate reactivity significantly predicted IES intrusion (ß = -0.208, t = -2.28, p = .025, ΔR2 = 0.041, confidence interval = -0.021 to -0.001) and IES hyperarousal (ß = -0.224, t = -2.54, p = .012, ΔR2 = 0.047, confidence interval = -0.22 to - 0.003), but not IES avoidance (p = .077). These results remained statistically significant after adjustment for sex, socioeconomic status, baseline cardiovascular activity, neuroticism, race, ethnicity, body mass index, and adverse childhood experiences. There were no statistically significant associations between blood pressure and any of the Impact of Event Scale-Revised subscales (p values > .12). CONCLUSIONS: Diminished heart rate responses (i.e., lower physiological arousal) to acute psychological stress before the COVID-19 pandemic significantly predicted reported PTSD symptoms during the crisis.

Individual differences in emotion regulation prospectively predict early COVID-19 related acute stress.

Preliminary prospective research suggests emotion dysregulation may confer vulnerability to poor stress responses. The present prospective study extends this research by examining both specific emotion regulation strategies and global emotion regulation difficulties in the context of acute stress following onset of the COVID-19 global pandemic in 119 young adults. As part of a larger study, emotion regulation was assessed prior to pandemic onset (January 2019 - February 2020) using two standard measures (Emotion Regulation Questionnaire, ERQ, Gross & John, 2003; Difficulties in Emotion Regulation Scale, DERS, Gratz & Roemer, 2004). A self-report assessment of acute stress was conducted 2-3½ weeks after the COVID-19 pandemic declaration. Results demonstrated cognitive reappraisal and expressive suppression (i.e., ERQ) were not individually predictive of acute stress; however, there was a significant interaction of suppression by reappraisal. Simple effects indicated suppression was negatively associated with acute stress only when reappraisal levels were high. Greater global emotion regulation difficulties (i.e., DERS), particularly nonacceptance of emotions and limited access to emotion regulation strategies, significantly predicted greater acute stress. These results provide further evidence of the temporal relationship between emotion dysregulation and stress reactions, and also suggest the expected effects of emotion regulation strategies may differ across contexts.

Pediatric Intern Clinical Exposure During the COVID-19 Pandemic.

BACKGROUND AND OBJECTIVES: Pediatric health care encounters declined during the coronavirus disease 2019 (COVID-19) pandemic, and pediatric residency programs have adapted trainee schedules to meet the needs of this changing clinical environment. We sought to evaluate the impact of the pandemic on pediatric interns' clinical exposure. METHODS: In this retrospective cohort study, we quantified patient exposure among pediatric interns from a single large pediatric residency program at a freestanding children's hospital. Patient encounters and shifts per pediatric intern in the inpatient and emergency department settings were evaluated during the COVID-19 pandemic, from March to June 2020, as compared with these 3 months in 2019. Patient encounters by diagnosis were also evaluated. RESULTS: The median number of patient encounters per intern per 2-week block declined on the pediatric hospital medicine service (37.5 vs 27.0; P < .001) and intensive care step-down unit (29.0 vs 18.8; P = .004) during the pandemic. No significant difference in emergency department encounters was observed (63.0 vs 40.5; P = .06). The median number of shifts worked per intern per 2-week block also decreased on the pediatric hospital medicine service (10.5 vs 9.5, P < .001). Across all settings, there were more encounters for screening for infectious disease and fewer encounters for respiratory illnesses. CONCLUSIONS: Pediatric interns at the onset of the COVID-19 pandemic were exposed to fewer patients and had reduced clinical schedules. Careful consideration is needed to track and supplement missed clinical experiences during the pandemic.

Emotion Regulation Strategies Predict PTSS During the COVID-19 Pandemic in an American Indian Population.

BACKGROUND: Poor emotion regulation is associated with post-traumatic stress symptoms (PTSS). However, limited prospective research prevents any directional conclusions. No known studies have assessed emotion regulation with PTSS in American Indians, a high-risk population for poor mental health outcomes. The present prospective study explored whether emotion regulation strategies (cognitive reappraisal, expressive suppression) predicted later PTSS related to the COVID-19 global pandemic in a solely American Indian sample. METHODS: American Indian participants (N = 210; Mean (SD) age = 54.85(13.08) years, 58.7% female) completed the Emotion Regulation Questionnaire (ERQ) during Phase 1 (a few weeks before pandemic declaration) and the Impact of Event Scale-Revised (IES-R) with respect to the COVID-19 pandemic during Phase 2 (7-8 weeks after pandemic declaration). Bivariate correlations and hierarchical linear regression analyses were utilized. RESULTS: ERQ reappraisal was negatively associated with IES-R total scores, such that higher reappraisal predicted lower PTSS. In contrast, ERQ suppression was positively associated with IES-R total scores, such that higher suppression predicted higher PTSS. CONCLUSIONS: Greater suppression and lower reappraisal predicts PTSS in response to the COVID-19 pandemic in an entirely American Indian sample, providing critical information for future interventions in a population at high-risk for mental health disparities.