Receiver Operator Characteristic (ROC) Curves

Nettleman, Mary D.

doi:10.2307/30145466

Cited by 22 publications

(12 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ROC characteristics describe the trade‐off between the selectivity and specificity of a clinical test (16). The analysis was initially developed from the early days of radar and sonar detection during the Second World War.…”

Section: Discussionmentioning

confidence: 99%

Prospective comparison of POSSUM and P‐POSSUM with clinical assessment of mortality following emergency surgery

Hobson

Sutton

Garcea

et al. 2006

Acta Anaesthesiol Scand

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

Prospective comparison of POSSUM and P‐POSSUM with clinical assessment of mortality following emergency surgery

Hobson

Sutton

Garcea

et al. 2006

Acta Anaesthesiol Scand

View full text Add to dashboard Cite

show abstract

“…The powerful analytical tools used in this study are broadly divided into univariate and multivariate. Univariate analytical tools included Mann–Whitney U test (non‐parametric statistical significance test used to calculate a confidence interval for the median difference between two independent groups of sampled data [Whitney, 1997]), receiver‐operator characteristic (ROC, plotting the dynamic relationship between sensitivity and specificity independent of disease prevalence [Nettleman, 1988]), scatter plot (displaying data points on a 2‐D graph), Box and Whisker plot (displaying a statistical summary of the data values: median, 25–75 percentile quartiles and minimum to maximum value range); whereas multivariate analytical tools included principal components analysis (PCA, reducing the complexity of data and distinguishing sample clusters [Jayakar, 1980]), heat map (presentation of a dendrogram representing similarities in the expression patterns between the sample groups), classification and regression tree analysis (CART, predicting multiple potential biomarkers and cross‐validation by applying the tree computed from learning sample to testing sample [Gribonval, 2005]).…”

Section: Methodsmentioning

confidence: 99%

Differential expression of proteins in kidney, eye, aorta, and serum of diabetic and non‐diabetic rats

Cho¹,

Yip²,

Chung³

et al. 2006

J of Cellular Biochemistry

View full text Add to dashboard Cite

Diabetes mellitus (DM) is a chronic progressive disease that often results in microvascular and macrovascular complications, yet its pathogenesis is not clear. Automated proteomic technology, coupled with powerful bioinformatics and statistical tools, can provide new insights into the molecular alterations implicated in DM. Following our previous findings of redox changes in the eye and aorta of diabetic rats, as well as the activities of different antioxidant enzymes during the development of DM, this study is further launched to find potential biomarkers by comparing the serum and tissue samples of 26 diabetic rats (8 weeks after streptozotocin [STZ] administration) with 29 normal controls using surface enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS) technology. Eight potential biomarkers were found in the serum, one potential biomarker was found in the kidney and eye, respectively, whereas three potential biomarkers were discovered in the aorta. One of the serum biomarker candidates was found to match the C-reactive protein (CRP) in the Swiss-Prot knowledgebase. Further validation has been conducted by ELISA kit to confirm the role of CRP during the development of DM. To conclude, the increased level of CRP in diabetic serum demonstrated in this study indicates that the development of DM is associated with inflammation. This is also the first report demonstrating that some potential lysate biomarkers in the kidney, eye, and aorta may be involved in the development of diabetes and its complications. Further identification and evaluation of these potential biomarkers will help unravel the underlying mechanisms of the disease.

show abstract

“…The graphic representation of this spectrum for a particular model is called the ROC curve. This curve is generated by changing the decision threshold over the entire range of possible prediction scores (in the present study from 0 to 5) and by calculating the sensitivity/specificity pairs of the decision matrix at each threshold 17,18 …”

Section: Methodsmentioning

confidence: 99%

“…The final multivariate proportional hazards regression model that was developed on the “development data set” was applied to the “validation data set” to observe its performance. The risk ratios, 95% CIs, area under the ROC curve (an estimate of the strength of the model), 17,19,20 and performance on the MRIS (estimation of the chance of dying in particular risk groups) were compared.…”

Section: Methodsmentioning

confidence: 99%