Browse Articles

Predictive modelling and identification of critical variables of mortality risk in COVID-19 patients

Abstract South Africa was the most affected country in Africa by the coronavirus disease 2019 (COVID-19) pandemic, where over 4 million confirmed cases of COVID-19 and over 102,000 deaths have been recorded since 2019. Aside from clinical methods, artificial intelligence (AI)-based solutions such as machine learning (ML) models have been employed in treating COVID-19 cases. However, limited application of AI for COVID-19 in Africa has been reported in the literature. This study aimed to investigate the performance and interpretability of several ML algorithms, including deep multilayer perceptron (Deep MLP), support vector machine (SVM) and Extreme gradient boosting trees (XGBoost) for predicting COVID-19 mortality risk with an emphasis on the effect of cross-validation (CV) and principal component analysis (PCA) on the results. For this purpose, a dataset with 154 features from 490 COVID-19 patients admitted into the intensive care unit (ICU) of Tygerberg Hospital in Cape Town, South Africa, during the first wave of COVID-19 in 2020 was retrospectively analysed. Our results show that Deep MLP had the best overall performance (F1 = 0.92; area under the curve (AUC) = 0.94) when CV and the synthetic minority oversampling technique (SMOTE) were applied without PCA. By using the Shapley Additive exPlanations (SHAP) model to interpret the mortality risk predictions, we identified the Length of stay (LOS) in the hospital, LOS in the ICU, Time to ICU from admission, days discharged alive or death, D-dimer (blood clotting factor), and blood pH as the six most critical variables for mortality risk prediction. Also, Age at admission, Pf ratio (PaO2/FiO2 ratio), troponin T (TropT), ferritin, ventilation, C-reactive protein (CRP), and symptoms of acute respiratory distress syndrome (ARDS) were associated with the severity and fatality of COVID-19 cases. The study reveals how ML could assist medical practitioners in making informed decisions on handling critically ill COVID-19 patients with comorbidities. It also offers insight into the combined effect of CV, PCA, and SMOTE on the performance of ML models for COVID-19 mortality risk prediction, which has been little explored.

Publisher Correction: A generic self-learning emotional framework for machines

Striking a Balance — Advancing Physician Collective-Bargaining Rights and Patient Protections

EEG-based functional and effective connectivity patterns during emotional episodes using graph theoretical analysis

Learning the fitness dynamics of pathogens from phylogenies

Abstract The dynamics of the genetic diversity of pathogens, including the emergence of lineages with increased fitness, is a foundational concept of disease ecology with key public-health implications. However, the identification of such lineages and estimation of associated fitness remain challenging, and is rarely done outside densely sampled systems 1,2 . Here we present phylowave, a scalable approach that summarizes changes in population composition in phylogenetic trees, enabling the automatic detection of lineages based on shared fitness and evolutionary relationships. We use our approach on a broad set of viruses and bacteria (SARS-CoV-2, influenza A subtype H3N2, Bordetella pertussis and Mycobacterium tuberculosis ), which include both well-studied and understudied threats to human health. We show that phylowave recovers the main known circulating lineages for each pathogen and that it can detect specific amino acid changes linked to fitness changes. Furthermore, phylowave identifies previously undetected lineages with increased fitness, including three co-circulating B.   pertussis lineages. Inference using phylowave is robust to uneven and limited observations. This widely applicable approach provides an avenue to monitor evolution in real time to support public-health action and explore fundamental drivers of pathogen fitness.

Sex, Gender, and Sexuality in the <i>Journal</i>

Identifying risk factors and constructing a predictive model for heart failure combined with intracardiac thrombus in non-compaction cardiomyopathy patients

KRAS Oncoprotein Signaling in Cancer

The segmentation of nanoparticles with a novel approach of HRU2-Net†

Abstract Nanoparticles have great potential for the application in new energy and aerospace fields. The distribution of nanoparticle sizes is a critical determinant of material properties and serves as a significant parameter in defining the characteristics of zero-dimensional nanomaterials. In this study, we proposed HRU2-Net†, an enhancement of the U2-Net† model, featuring multi-level semantic information fusion. This approach exhibits strong competitiveness and refined segmentation capabilities for nanoparticle segmentation. It achieves a Mean intersection over union (MIoU) of 87.31%, with an accuracy rate exceeding 97.31%, leading to a significant improvement in segmentation effectiveness and precision. The results show that the deep learning-based method significantly enhances the efficacy of nanomaterial research, which holds substantial significance for the advancement of nanomaterial science.

Influenza Vaccination Strategies in Patients with Hematologic Cancer

Daily walking habits can mitigate age-related decline in static balance: a longitudinal study among aircraft assemblers

Abstract Improving physical balance among older workers is essential for preventing falls in workplace. We aimed to elucidate the age-related decline in one-leg standing time with eyes closed, an indicator of static balance, and mitigating influence of daily walking habits on this decline in Japan. This longitudinal study involved 249 manufacturing workers, including seven females, aged 20–66 years engaged in tasks performed at height in the aircraft and spacecraft machinery industry. The participants underwent a one-leg standing test and annual health checkups through the Kanagawa Health Service Association between 2017 and 2019. The outcome measure was one-leg standing time up to 30 s. The coefficient (β) of one-leg standing time against aging was estimated using two-level multilevel linear regression with random intercepts. We also estimated the β of daily walking habits at least one hour per day. The quadratic spline curve showed an almost linear trend of one-leg standing time with age. The one-leg standing time significantly decreased with age (adjusted β = − 0.22; 95% confidence interval [CI] − 0.31 to − 0.14). Meanwhile, walking habits showed a preventive effect (β = 1.76; 95% CI 0.49 to 3.04). Age-related decline in one-leg standing time may be mitigated by simple daily walking habits.

Inavolisib Therapy in Advanced Breast Cancer

Treatment of patients with tumor/treatment-related hypothalamic obesity in the first two years following surgical treatment or radiotherapy

Abstract Survivors of sellar/suprasellar tumors involving hypothalamic structures face a risk of impaired quality of life, including tumor- and/or treatment-related hypothalamic obesity (TTR-HO) defined as abnormal weight gain resulting in severe persistent obesity due to physical, tumor- and/or treatment related damage of the hypothalamus. We analyze German claims data to better understand treatment pathways for patients living TTR-HO during the two years following the index surgical treatment. A database algorithm identified patients with TTR-HO in a representative German payer claims database between 2010 and 2021 (n = 5.42 million patients). Claims from 37 patients with TTR-HO were analyzed on a quarterly basis over 2 years. The analysis considered inpatient, outpatient, and prescription data. In the follow-up period, patients with TTR-HO are hospitalized 3.68 times on average; 37% of hospitalizations in year 1 and 31% in year 2 are due to TTR-HO. On average, patients see a general practitioner 12.27 times and various specialists 20.45 times. The need for complex neuroendocrine therapy develops quickly, with most patients having 2–3 neuroendocrine prescriptions in any given quarter. The management of patients with TTR-HO requires frequent inpatient and outpatient visits for tumor follow-up and management of incident comorbidities, and most patients with TTR-HO require intense polytherapy.

Survival with Trastuzumab Emtansine in Residual HER2-Positive Breast Cancer

An in-depth study of indolone derivatives as potential lung cancer treatment

Infrequent Zoledronate — Small Individual Gain, Larger Population Gain

Sex disparities in the association between rare earth elements exposure and genetic mutation frequencies in lung cancer patients

Abstract The ubiquitous use of rare earth elements (REEs) in modern living environments raised concern about their impact on human health. With the detrimental and beneficial effects of REEs reported by different studies, the genuine role of REEs in the human body remains a mystery. This study explored the association between REEs and genetic mutations in patients with lung adenocarcinoma (LUAD). A cohort of 53 LUAD patients underwent tumor DNA sequencing (1123 cancer-related genes) and plasma REE (lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), and yttrium (Y)) quantification. We found divergent relationships between plasma REE levels and mutation load between sexes. Specifically, Ce levels and mutation load were positively correlated in males but negatively correlated in females, while La exposure exhibited opposite associations in the two sexes. This observation was validated using the Bayesian Kernel Machine Regression (BKMR) model. Additionally, plasma REE levels was associated with specific mutation types and variant allele frequencies (VAFs) of particular genes in a sex-dependent manner. Mutational signature analysis revealed sex-specific associations of La with indel signatures. These findings highlight the intricate interplay between plasma REE levels and genetic mutations in LUAD, emphasizing the need for a personalized, sex-oriented approach to understand and treat this disease.

Asymptomatic Severe Aortic-Valve Stenosis — To Wait or Not to Wait

Metabolomics identifies plasma biomarkers of localized radiation injury

Abstract A radiological accident may result in the development of a local skin radiation injury (LRI) which may evolve, depending on the dose, from dry desquamation to deep ulceration and necrosis through unpredictable inflammatory waves. Therefore, early diagnosis of victims of LRI is crucial for improving medical care efficiency. This preclinical study aims to identify circulating metabolites as biomarkers associated with LRI using a C57BL/6J mouse model of hind limb irradiation. More precisely, two independent mice cohorts were used to conduct a broad-spectrum profiling study followed by a suspect screening analysis performed on plasma metabolites by mass spectrometry. An integrative analysis was conducted through a multi-block sparse partial least square discriminant analysis (sPLS-DA) to establish multi-scale correlations between specific metabolites levels and biological, physiological (injury severity), and functional parameters (skin perfusion). The identified biomarker signature consists in a 6-metabolite panel including putrescine, uracil, 2,3-dihydroxybenzoate, 3-hydroxybenzoate, L-alanine and pyroglutamate, that can discriminate mice according to radiation dose and injury severity. Our results demonstrate relevant molecular signature associated with LRI in mice and support the use of plasma metabolites as suitable molecular biomarkers for LRI prognosis and diagnosis.

Exogenous Lipoid Pneumonia