Bayesian feature selection for radiomics using reliability metrics.

Introduction: Imaging of tumors is a standard step in diagnosing cancer and making subsequent treatment decisions. The field of radiomics aims to develop imaging based biomarkers using methods rooted in artificial intelligence applied to medical imaging. However, a challenging aspect of developing predictive models for clinical use is that many quantitative features derived from image data exhibit instability or lack of reproducibility across different imaging systems or image-processing pipelines. Methods: To address this challenge, we propose a Bayesian sparse modeling approach for image classification based on radiomic features, where the inclusion of more reliable features is favored via a probit prior formulation. Results: We verify through simulation studies that this approach can improve feature selection and prediction given correct prior information. Finally, we illustrate the method with an application to the classification of head and neck cancer patients by human papillomavirus status, using as our prior information a reliability metric quantifying feature stability across different imaging systems.

Themes Depicted in Top-grossing Rated-R Films Released from 2012 to 2017.

INTRODUCTION: Films continue to be a popular form of entertainment amongst children and adolescents. The objective of this study was to determine the most common themes depicted in a select number of rated-R films released 2012-2017. METHODS: A total of 25 films were independently viewed and analyzed by five reviewers. RESULTS: The most common positive themes were "encouragement", "helping others", "teamwork", "compassion", and "honesty" (2.3, 1.7, 1.7, 1.6, 1.4 mean events per hour, respectively). The most common negative themes were "use of inappropriate language", "use of a lethal weapon", "physical fighting", "dishonesty", and "demonstrating excessive anger" (8.9, 4.4, 3.4, 2.3, 2.2 mean events per hour, respectively). CONCLUSION: The most common positive themes in our sample were associated with service, collaboration, and honesty/humility, and the most common negative themes were associated with violence. We encourage co-viewing and active mediation, focusing on the themes found in films, as a method to guide children through their development process.

The impact of patient-physician alliance on trust following an adverse event.

OBJECTIVE: Adverse events in maternity care have a negative impact on the patient-physician relationship. This study assesses the effects of healthcare institutions, communication, and patient involvement on patient trust following adverse events. METHODS: Surveys were distributed online to women across the US who had given birth. Women were asked to recount their experiences during their most recent pregnancy including unexpected procedures, adverse events, support from healthcare institutions, and perceived betrayals by healthcare institutions. RESULTS: Adverse events were negatively correlated with patient trust in their physician. This study's results illustrated that patient involvement and institutional betrayal mediated patient trust following adverse events. Patients who were more involved in decision-making with their physician were found to have more trust in them following adverse events than those who did not. CONCLUSIONS: Patient-physician trust is negatively affected by adverse events, but patient-physician alliance in decision-making can decrease this impact. Therefore, physicians can work proactively to lessen the detrimental effects of adverse events on patient trust, but the patient-physician relationship is still impacted by healthcare institutions. PRACTICE IMPLICATIONS: This study demonstrates the benefits of encouraging patient alliance with their physician and supports a need for education on the use of these strategies in healthcare.

Hierarchical Normalized Completely Random Measures for Robust Graphical Modeling.

Gaussian graphical models are useful tools for exploring network structures in multivariate normal data. In this paper we are interested in situations where data show departures from Gaussianity, therefore requiring alternative modeling distributions. The multivariate t-distribution, obtained by dividing each component of the data vector by a gamma random variable, is a straightforward generalization to accommodate deviations from normality such as heavy tails. Since different groups of variables may be contaminated to a different extent, Finegold and Drton (2014) introduced the Dirichlet t-distribution, where the divisors are clustered using a Dirichlet process. In this work, we consider a more general class of nonparametric distributions as the prior on the divisor terms, namely the class of normalized completely random measures (NormCRMs). To improve the effectiveness of the clustering, we propose modeling the dependence among the divisors through a nonparametric hierarchical structure, which allows for the sharing of parameters across the samples in the data set. This desirable feature enables us to cluster together different components of multivariate data in a parsimonious way. We demonstrate through simulations that this approach provides accurate graphical model inference, and apply it to a case study examining the dependence structure in radiomics data derived from The Cancer Imaging Atlas.

Tree-based Methods for Characterizing Tumor Density Heterogeneity.

Solid lesions emerge within diverse tissue environments making their characterization and diagnosis a challenge. With the advent of cancer radiomics, a variety of techniques have been developed to transform images into quantifiable feature sets producing summary statistics that describe the morphology and texture of solid masses. Relying on empirical distribution summaries as well as grey-level co-occurrence statistics, several approaches have been devised to characterize tissue density heterogeneity. This article proposes a novel decision-tree based approach which quantifies the tissue density heterogeneity of a given lesion through its resultant distribution of tree-structured dissimilarity metrics computed with least common ancestor trees under repeated pixel re-sampling. The methodology, based on statistics derived from Galton-Watson trees, produces metrics that are minimally correlated with existing features, adding new information to the feature space and improving quantitative characterization of the extent to which a CT image conveys heterogeneous density distribution. We demonstrate its practical application through a diagnostic study of adrenal lesions. Integrating the proposed with existing features identifies classifiers of three important lesion types; malignant from benign (AUC = 0.78), functioning from non-functioning (AUC = 0.93) and calcified from non-calcified (AUC of 1).

Childhood asthma and extreme values of body mass index: the Harlem Children's Zone Asthma Initiative.

To examine the association between body mass index (BMI) percentile and asthma in children 2-11 years of age, we performed a cross-sectional analysis of 853 Black and Hispanic children from a community-based sample of 2- to 11-year olds with measured heights and weights screened for asthma by the Harlem Children's Zone Asthma Initiative. Current asthma was defined as parent/guardian-reported diagnosis of asthma and asthma-related symptoms or emergency care in the previous 12 months. Among girls, asthma prevalence increased approximately linearly with increasing body mass index (BMI) percentile, from a low of 12.0% among underweight girls (BMI 95th percentile). After adjusting for age, race/ethnicity, and household smoking, among girls, having asthma was associated with being at risk for overweight (odds ratio [OR], 2.6; 95% confidence interval [CI], 1.4-5.0) and being overweight (OR, 2.1; 95% CI, 1.2-3.8) compared to normal weight; among boys, having asthma was associated both with overweight (OR, 2.4; 95% CI, 1.4-4.3) and with underweight (OR, 2.9; 95% CI, 1.1-7.7). Large, prospective studies that include very young children are needed to further explore the observed association between underweight and asthma among boys. Early interventions that concomitantly address asthma and weight gain are needed among pre-school and school-aged children.

Addressing the childhood asthma crisis in Harlem: the Harlem Children's Zone Asthma Initiative.

OBJECTIVES: We determined the prevalence of asthma and estimated baseline asthma symptoms and asthma management strategies among children aged 0-12 years in Central Harlem. METHODS: The Harlem Children's Zone Asthma Initiative is a longitudinal, community-based intervention designed for poor children with asthma. Children aged 0-12 years who live or go to school in the Harlem Children's Zone Project or who participate in any Harlem Children's Zone, Inc, program were screened for asthma. Children with asthma or asthma-like symptoms were invited to participate in an intensive intervention. RESULTS: Of the 1982 children currently screened, 28.5% have been told by a doctor or nurse that they have asthma, and 30.3% have asthma or asthma-like symptoms. To date, 229 children are enrolled in the Harlem Children's Zone Asthma Initiative; at baseline, 24.0% had missed school in the last 14 days because of asthma. CONCLUSION: The high prevalence of asthma among children in the Harlem Children's Zone Project is consistent with reports from other poor urban communities. Intensive efforts are under way to reduce children's asthma symptoms and improve their asthma management strategies.