Individual ROC Research Articles

A comparative proteomic analysis for non-invasive early prediction of hypoxic-ischemic injury in asphyxiated neonates.

Hypoxic Ischemic Encephalopathy (HIE) is one of the principal causes of neonatal mortality and long-term morbidity worldwide. The neonatal signs of mild cerebral injury are subtle, making an early precise diagnosis difficult. Delayed detection, poor prognosis, and lack of specific biomarkers for the disease are increasing mortality rates. In this study, we intended to identify specific biomarkers using comparative proteomic analysis to predict the severity of perinatal asphyxia so that its outcome can also be prevented. A case-control study was conducted on 38 neonates, and urine samples were collected within 24 and 72h of life. A tandem mass spectrometry-based quantitative proteomics approach, followed by validation via sandwich ELISA, was performed. The LC-MS/MS-based proteomics analysis resulted in the identification of 1201 proteins in urine, with 229, 244, and 426 being differentially expressed in HIE-1, HIE-2, and HIE-3, respectively. Axon guidance, Diseases of programmed cell death, and Detoxification of reactive oxygen species pathways were significantly enriched in mild HIE versus severe HIE. Among the differentially expressed proteins in various stages of HIE, we chose to validate four proteins - APP, AGT, FABP1, and FN1 - via sandwich ELISA. Individual and cumulative ROC curves were plotted. AGT and FABP1 together showed high sensitivity, specificity, and accuracy as potential biomarkers for early diagnosis of HIE. Establishing putative urinary biomarkers will facilitate clinicians to more accurately screen neonates for brain injury and monitor the disease progression. Prompt treatment of neonates may reduce mortality and neurodevelopmental impairment.

Development and validation of multivariable clinical diagnostic models to identify type 1 diabetes requiring rapid insulin therapy in adults aged 18–50 years

ObjectiveTo develop and validate multivariable clinical diagnostic models to assist distinguishing between type 1 and type 2 diabetes in adults aged 18–50.DesignMultivariable logistic regression analysis was used to develop classification...

Towards a PBMC “virogram assay” for precision medicine: Concordance between ex vivo and in vivo viral infection transcriptomes

BackgroundUnderstanding individual patient host-response to viruses is key to designing optimal personalized therapy. Unsurprisingly, in vivo human experimentation to understand individualized dynamic response of the transcriptome to viruses are rarely studied because of the obvious limitations stemming from ethical considerations of the clinical risk. ObjectiveIn this rhinovirus study, we first hypothesized that ex vivo human cells response to virus can serve as a proxy for otherwise controversial in vivo human experimentation. We further hypothesized that the N-of-1-pathways framework, previously validated in cancer, can be effective in understanding the more subtle individual transcriptomic response to viral infection. MethodN-of-1-pathways computes a significance score for a given list of gene sets at the patient level, using merely the ‘omics profiles of two paired samples as input. We extracted the peripheral blood mononuclear cells (PBMC) of four human subjects, aliquoted in two paired samples, one subjected to ex vivo rhinovirus infection. Their dysregulated genes and pathways were then compared to those of 9 human subjects prior and after intranasal inoculation in vivo with rhinovirus. Additionally, we developed the Similarity Venn Diagram, a novel visualization method that goes beyond conventional overlap to show the similarity between two sets of qualitative measures. ResultsWe evaluated the individual N-of-1-pathways results using two established cohort-based methods: GSEA and enrichment of differentially expressed genes. Similarity Venn Diagrams and individual patient ROC curves illustrate and quantify that the in vivo dysregulation is recapitulated ex vivo both at the gene and pathway level (p-values⩽0.004). ConclusionWe established the first evidence that an interpretable dynamic transcriptome metric, conducted as an ex vivo assays for a single subject, has the potential to predict individualized response to infectious disease without the clinical risks otherwise associated to in vivo challenges. These results serve as a foundational work for personalized “virograms”.

Osteopontin: a novel predictor of survival in patients with systemic light-chain amyloidosis

Background: Troponin-T (cTnT) and NT-proBNP provide prognostic information in light-chain amyloidosis (AL). Thus, these biomarkers are widely used in clinical routine for risk stratification. Recently, plasma level of osteopontin (OPN), a secreted phosphoglycoprotein expressed by a variety of cell types, has been reported as a risk predictor in various cardiovascular diseases.Methods: OPN was determined retrospectively in 150 consecutive patients newly diagnosed with AL amyloidosis. All patients were evaluated according to a routine protocol including electrocardiography, echocardiography and laboratory testing.Results: Mean OPN was 591 ± 37 ng/mL. Cardiac involvement was established in 83 (55.3%). Median OPN plasma level were associated with number of organs involved, renal function, eligibility for high-dose melphalan chemotherapy and autologous stem cell transplantation, and severity of cardiac amyloidosis. Median follow-up was 19.2 months. 1-year all-cause-survival was 83.4%. The cut-offs discriminating 1-year all-cause-mortality for NT-proBNP, troponin T, and OPN were 2544 ng/L, 0.035 µg/L, and 426.8 ng/mL, respectively. Outcome was worse in patients with biomarkers above the individual ROC derived cut-off. A significant improvement of survival was observed in patients with cTNT >0.035 µg/L or NT-proBNP >2544 ng/L and OPN below ROC-derived cut-off of 426.8 ng/mL as compared to patients with OPN above 426.8 ng/L. No further discrimination was achieved by OPN in the cohorts of low troponin T or low NT-proBNP, respectively. Separate multivariate models identified OPN (cut-off 426.8 ng/mL) and troponin T (cut-off 0.035 µg/L) as independent predictors of all-cause-mortality.Conclusions: These data demonstrated that OPN appears to be a valuable marker in the clinical routine for evaluation of patients with AL amyloidosis, especially if it is used in combination with cTNT and/or NT-proBNP.

Comparison of Semiparametric, Parametric, and Nonparametric ROC Analysis for Continuous Diagnostic Tests Using a Simulation Study and Acute Coronary Syndrome Data

We aimed to compare the performance of three different individual ROC methods (one from each of the broad categories of parametric, nonparametric and semiparametric analysis) for assessing continuous diagnostic tests: the binormal method as a parametric method, an empirical approach as a nonparametric method, and a semiparametric method using generalized linear models (GLM). We performed a simulation study with various sample sizes under normal, skewed, and monotone distributions. In the simulations, we used estimates of the ROC curve parameters a and b, estimates of the area under the curve (AUC), the standard errors and root mean square errors (RMSEs) of these estimates, and the 95% AUC confidence intervals for comparison. The three methodologies were also applied to an acute coronary syndrome dataset in which serum myoglobin levels were used as a biomarker for detecting acute coronary syndrome. The simulation and application studies suggest that the semiparametric ROC analysis using GLM is a reliable method when the distributions of the diagnostic test results are skewed and that it provides a smooth ROC curve for obtaining a unique cutoff value. A sample size of 50 is sufficient for applying the semiparametric ROC method.

Conventional and non-conventional coronary risk factors in male premature coronary artery disease patients already having a low Framingham risk score

Background — Individual risk factors and, more importantly, global risk assessment tools such as the Framingham risk score have been used successfully for risk prediction especially in older patients. However, there is paucity of data about the coronary heart disease prediction in premature coronary artery disease patients with a low Framingham risk score.Methods and results — We recruited 102 consecutive young patients without hypertension and diabetes mellitus in the study. All subjects had had chest pain and underwent coronary angiography since non-invasive diagnostic test results suggested ischaemia. Forty-five patients having at least one coronary lesion independent of severity were included in the study group.The remaining fifty-seven subjects without any coronary lesion were used as control group. Conventional and non-conventional risk factors were evaluated both in patients and control subjects. Framingham risk score and absolute 10-year hard CHD events risk were also calculated for each individual.The coronary heart disease group had a significantly higher smoking frequency as compared to the control group. They also had higher plasma levels of triglycerides, apolipoprotein B and apo B/A1 ratio but a smaller LDL particle size.We failed to find any independent CHD predictor after logistic regression analysis. However, individual ROC curve analysis of risk factors revealed that apolipoprotein B, triglycerides and apo B/A1 ratio have the highest area under the curve for coronary artery disease prediction.Conclusions — The Framingham risk score may underestimate the true risk of an individual. Incorporating non-conventional risk factors such as apolipoprotein B and apo B/apo A1 ratio may provide valuable information in these patients.

Modeling boolean decision rules applied to multiple-observer decision strategies.

A model that derives multiple-observer decision strategy ROC curves for boolean decision rules applied to binary decisions of two or three observers is presented. It is assumed that covert decision variables consistent with ROC models of observer performance underlie decisions and that readers' decision criteria are in a fixed relationship. The specific parameters of individual ROC curves and the correlational structure that describes interobserver agreement have dramatic effects upon the relative benefits to be derived from different boolean strategies. A common strategy employed in clinical practice, in which the overall decision is positive if any observer makes a positive decision, is most effective when the readers are of similar ability, when they adopt similar decision criteria, when interreader agreement is greater for negative than for positive cases, and when the individual ROC slope is <<1.0. Different multiple-observer decision strategies can be evaluated using the model equations. A bootstrap method for testing model-associated hypotheses is described.

Experiences with a Teleradiology System in Pulmonary Diseases

Chest radiographs of 43 patients were digitized and sent from Paimio Hospital to Turku University Central Hospital. Fifteen of the patients had a pneumothorax, 12 had interstitial lung disease and 16 were controls. The images were interpreted by 5 radiologists during their duty shift. The first reading session consisted of images with 1,024 x 1,024 pixel resolution and the second of images with 2,048 x 2,048 matrix followed by the original chest radiographs. ROC studies were performed from the answers based on a 5-point confidence scale. In the pneumothorax group the average area under the individual ROC curve was 0.928 with the 1,024 x 1,024 matrix and 0.983 with the 2,048 x 2,048 matrix. In the fibrosis group the average area under the individual ROC curve was 0.877 with the 1,024 x 1,024 matrix and 0.831 with the 2,048 x 2,048 matrix. The results suggest that the 1,024 x 1,024 matrix is adequate in the detection of pneumothoraces and fibrosis.

The effects of memory load and delay on facial recognition

AbstractThis experiment examined the effects of memory load over three periods of delay. Following presentation of 20, 35, or 50 targets, subjects were required to select these from an equal number of distractors 10 min, 1 week, or 2 weeks later. Increased target load (independent of increases in recognition load) decreased accuracy mainly by decreasing hit rate. Increasing delay decreased accuracy largely as a result of an increased false alarm rate. Most individual ROC curves, plotted on z‐coordinates, had slopes &lt; 1. Implications of this finding for the use of d' are discussed.

External quality control of gamma cameras —results of an inter-laboratory comparison study in Austria

Some 28 nuclear medicine departments out of a total of 40 participated in a nationwide Austrian interlaboratory comparison of the imaging quality of gamma cameras. The participation was voluntary, and confidentiality of the individual results was guaranteed. The survey was completed within 2 months, employing instructors to administrate the tests. An emission phantom simulating a flood field with non-uniformities was imaged by 43 cameras, and 54 images were evaluated. The test images were read by the participants using a graded rating scale to indicate the probability of the presence of a non-uniformity in the various parts of the image. The rating data were used to construct individual ROC curves for each image. The area under the ROC curve was used as the ranking parameter for image quality. The results show a spread of the ROC areas between 0.6 and 0.99, with a median of 0.81. A correlation was found between the year of installation and the ROC area obtained from the gamma camera, indicating improvements of performance in more recent cameras, which accounts for part of the variation of the ROC areas. The remaining variations are due to differences in the performance of the gamma cameras. Feedback was provided to the participants by describing individual performance with respect to the true structure of the phantom and by comparing this performance with that of the group. A questionnaire accompanying the test phantom yielded information about the practice of routine quality control and about details of the acquisition and analysis of images.(ABSTRACT TRUNCATED AT 250 WORDS)

Broadbent and Gregory Revisited: Vigilance and Statistical Decision

By rigorous test, Broadbent and Gregory (1963) showed that the behavior of subjects in a vigilance task could be adequately described by the equal variance model of signal detection theory (SDT). Since then the model has been widely applied in the analysis and interpretation of numerous vigilance studies, but its goodness of fit to the data has been met by mere assumption. The present study attempted to redress the balance by testing the assumption in respect of some 200 individual ROCs obtained from four vigilance experiments. Although the analysis indicated that about half of the individual ROC functions were of the curved form predicted by the equal-variance SDT model and hence support the view that vigilance performance can be interpreted in terms of this theoretical model of decision-making, a substantial proportion (30%) of the individual functions were of a particular form whose presence is difficult to account for, if handled within the SDT framework. Other models which seem more appropriate are suggested and the implications discussed.