Receiver Operating Characteristic

Workplace bullying is a serious hazard in every day working life that needs to be assessed carefully. In order to achieve such a goal, both victims of severe bullying as well as targets of less intensive bullying need to be identified, the latter in order to prevent further escalation into severe bullying. Previous research has mainly offered simple and crude measures of who is and who is not a victim of bullying. In this article we show how cutoff scores for the scale Negative Acts Questionnaire–Revised (NAQ-R) can be calculated. Based on a representative sample of the Norwegian workforce, we formulated 2 cutoff points or thresholds for the NAQ-R with a Receiver Operation Characteristic curve. Employees with a score lower than 33 are not bullied, employees with a score between 33 and 45 may be considered as being bullied occasionally, and employees who score above 45 can be considered to be victims of workplace bullying. As the NAQ-R is used in more than 300 research projects world...

Background: The double-blind, placebo-controlled food challenge (DBPCFC) is the “gold standard” for diagnosis of food hypersensitivity. Skin prick tests and RASTs are sensitive indicators of food-specific IgE antibodies but poor predictors of clinical reactivity. Previous studies suggested that high concentrations of food-specific IgE antibody were predictive of food-induced clinical symptoms. Because the CAP System FEIA (Pharmacia Diagnostics, Uppsala, Sweden) provides a quantitative assessment of allergen-specific IgE antibody, this study was undertaken to determine the potential utility of the CAP System FEIA in diagnosis of IgE-mediated food hypersensitivity. Methods: Sera from 196 patients with food allergy were analyzed for specific IgE antibodies to egg, milk, peanut, soy, wheat, and fish by CAP System FEIA. Sera were randomly selected from 300 stored samples of children and adolescents who had been evaluated by history, skin prick tests, and DBPCFCs. The study population was highly atopic; all patients had atopic dermatitis, and approximately 50% had asthma and allergic rhinitis at the time of initial evaluation. The performance characteristics of the CAP System FEIA were compared with those of skin prick tests and the outcome of DBPCFCs or “convincing” histories of anaphylactic reactions. Results: The prevalence of specific food allergies in the study population varied from 22% for wheat to 73% for egg. Allergy to egg, milk, peanut, and soy accounted for 87% of confirmed reactions. The performance characteristics of skin prick tests and CAP System FEIA (egg, milk, peanut, fish) were comparable, with excellent sensitivity and negative predictive accuracy but poor specificity and positive predictive accuracy. The performance characteristics of the CAP System FEIA for soy and wheat were poor. For egg, milk, peanut, and fish allergy, diagnostic levels of IgE, which could predict clinical reactivity in this population with greater than 95% certainty, were identified: egg, 6 kilounits of allergen-specific IgE per liter (kUA/L); milk, 32 kUA/L; peanut, 15 kUA/L; and fish, 20 kUA/L. Conclusions: When compared with the outcome of DBPCFCs, results of CAP System FEIA are generally comparable to those of skin prick tests in predicting symptomatic food hypersensitivity. Furthermore, by measuring the concentrations of food-specific IgE antibodies with the CAP System FEIA, it is possible to identify a subset of patients who are highly likely (>95%) to experience clinical reactions to egg, milk, peanut, or fish. This could eliminate the need to perform DBPCFCs in a significant number of patients suspected of having IgE-mediated food allergy. (J Allergy Clin Immunol 1997;100:444-51.)

Successful weaning from prolonged mechanical ventilation (MV) is an important issue in respiratory care centers (RCCs). Delayed or premature extubation increases both the risk of adverse outcomes and healthcare costs. However, the accurate evaluation of the timing of successful weaning from MV is very challenging in RCCs. This study aims to utilize artificial intelligence algorithms to build predictive models for the successful timing of the weaning of patients from MV in RCCs and to implement a dashboard with the best model in RCC settings. A total of 670 intubated patients in the RCC in Chi Mei Medical Center were included in the study. Twenty-six feature variables were selected to build the predictive models with artificial intelligence (AI)/machine-learning (ML) algorithms. An interactive dashboard with the best model was developed and deployed. A preliminary impact analysis was then conducted. Our results showed that all seven predictive models had a high area under the receive...

Data analysis is widely used to generate new insights into human disease mechanisms and provide better treatment methods. In this work, we used the mechanistic models of viral infection to generate synthetic data of influenza and COVID-19 patients. We then developed and validated a supervised machine learning model that can distinguish between the two infections. Influenza and COVID-19 are contagious respiratory illnesses that are caused by different pathogenic viruses but appeared with similar initial presentations. While having the same primary signs COVID-19 can produce more severe symptoms, illnesses, and higher mortality. The predictive model performance was externally evaluated by the ROC AUC metric (area under the receiver operating characteristic curve) on 100 virtual patients from each cohort and was able to achieve at least AUC=91% using our multiclass classifier. The current investigation highlighted the ability of machine learning models to accurately identify two differ...

Evaluation metric plays a critical role in achieving the optimal classifier during the classification training. Thus, a selection of suitable evaluation metric is an important key for discriminating and obtaining the optimal classifier. This paper systematically reviewed the related evaluation metrics that are specifically designed as a discriminator for optimizing generative classifier. Generally, many generative classifiers employ accuracy as a measure to discriminate the optimal solution during the classification training. However, the accuracy has several weaknesses which are less distinctiveness, less discriminability, less informativeness and bias to majority class data. This paper also briefly discusses other metrics that are specifically designed for discriminating the optimal solution. The shortcomings of these alternative metrics are also discussed. Finally, this paper suggests five important aspects that must be taken into consideration in constructing a new discriminator metric.

To identify factors predicting improvement/stabilization on the Alzheimer's Disease Cooperative Study-Clinical Global Impression of Change (ADCS-CGIC) and investigate whether early treatment responses can predict long-term outcomes, during a trial of 13.3 mg/24 h versus 4.6 mg/24 h rivastigmine patch in patients with severe Alzheimer's disease (AD). Logistic regression was used to relate Week 24 ADCS-CGIC score to potential baseline predictors. Additional analyses based on receiver-operating characteristic curves were performed using Week 8/16 ADCS-CGIC scores to predict response (13.3 mg/24 h patch) at Week 24. ADCS-CGIC score of (1) 1-3 = "improvement," (2) 1-4 = "improvement or no change". "Treatment" (13.3 mg/24 h patch) and increased age were significant predictors of "improvement" (P = 0.01 and P = 0.003, respectively), and "treatment" (P = 0.001), increased age (P = 0.002), and prior AD treatment (P = 0.03) for "i...

The detection rates of small lung cancers, especially adenocarcinoma, have recently increased. An understanding of malignant aggressiveness is critical for the selection of suitable therapeutic strategies, such as sublobar resection. The objective of this study was to examine the malignant biological behavior of clinical stage IA adenocarcinoma and to select therapeutic strategies using high-resolution computed tomography, fluorodeoxyglucose-positron emission tomography/computed tomography, and a pathologic analysis in the setting of a multicenter study.

The present study aims to evaluate the serum level of copper (Cu) in lambs suffering from subclinical forms of white muscle disease (WMD) and its relationship with cardiac troponin I (cTn-I) as a novel biomarker of cardiovascular disorders. Ten milliliters of jugular blood were taken from 200 lambs less than one year old to measure serum concentrations of Cu, selenium (Se), and cTn-I. The subjects were divided into 2 groups, namely, the deficient group which included 36 lambs, and the control group which included 164 lambs according to the reference serum Se concentration (50 ng/mL). Serum Se levels in the deficient group were lower than 50 ng/mL. By contrast, the control group showed Se levels higher than 50 ng/mL. Differences among the serum Cu and cTn-I levels were determined in both groups. The mean ±SD and median of serum Cu and cTn-I levels in the deficient group were lower and higher than those in the control group, respectively. A significant positive correlation was observed between serum Cu and Se levels, and also serum Cu and Se levels showed a negative correlation with serum cTn-I concentrations. Stepwise linear regression analysis showed that serum Cu levels were correlated positively with serum Se levels (p<0.05). Receiver operating characteristic (ROC) curve analysis indicated that the area under curve (AUC) of Cu was significantly higher than that of cTn-I based on the reference diagonal line. It is important to keep in mind that the value of AUC for the ROC curve is between 0.5 and 1.00, in which the lowest accuracy is related to the reference diagonal line with AUC of 0.5. A cut-off was determined to indicate which Cu level can discriminate between affected and healthy lambs. The cut-off level, sensitivity, and specificity of Cu in this study were 144.5 ng/mL, 74%, and 61%, respectively. © 2013 Ataollahi et al.

http://www.plosone.org/article/fetchObject.action?uri=info%3Adoi%2F10.1371%2Fjournal.pone.0056163&representation=PDF 
dx.plos.org/10.1371/journal.pone.0056163.g002

AIM Examination of psychometric properties and diagnostic accuracy of the Depression Self-Rating Scale for Adolescents (DSRS-A) as well as development and evaluations of a shorter version, DSRS-A-Screener. METHODS Analyses of component structure and internal consistency were performed in a community-based sample of adolescents N = 4,506 and among consecutive outpatients from three child psychiatric settings in Sweden (n = 137). Concurrent validity was measured as a correlation between a summation index of the scale items and the total major depressive disorder (MDD) symptom severity score from the Kiddie Schedule of Affective Disorders and Schizophrenia (K-SADS). Diagnostic accuracy was examined in the clinical sample, with the K-SADS interview as the reference test, by receiver operating characteristic analysis (ROC), calculations of sensitivity, specificity among other measures. With the purpose to select items for a shorter scale, associations between scale items and MDD were examined with binary logistic regression. This shorter scale was thereafter examined similarly. RESULTS Based on association with MDD, five items were selected for the brief DSRS-A Screener that showed one component structure, internal consistency Cronbach&#39;s alpha .80 and.82, respectively. In the clinical population concurrent validity was Spearman&#39;s rho .63 and ROC analysis showed AUC .84 (95% CI .78-.91; p &lt; .001). The optimal cut-off for screening was 2 with sensitivity .85 and specificity of .64. CONCLUSION The DSRS-A Screener compared to the original scale, maintained or improved reliability, validity, and showed moderate diagnostic accuracy.

PURPOSE Allogenic hematopoietic stem-cell transplant (HCT) is a curative therapy for acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). Relapse post-HCT is the most common cause of treatment failure and is associated with a poor prognosis. Pathologist-based visual assessment of aspirate images and the manual myeloblast counting have shown to be predictive of relapse post-HCT. However, this approach is time-intensive and subjective. The premise of this study was to explore whether computer-extracted morphology and texture features from myeloblasts&#39; chromatin patterns could help predict relapse and prognosticate relapse-free survival (RFS) after HCT. MATERIALS AND METHODS In this study, Wright-Giemsa–stained post-HCT aspirate images were collected from 92 patients with AML/MDS who were randomly assigned into a training set ( S t = 52) and a validation set ( S v = 40). First, a deep learning–based model was developed to segment myeloblasts. A total of 214 texture and ...