Statistical Study to Classify ?-Thalassaemia diseases in Erbil City at Thalassaemia Center by Using ROC Curve Analysis

Aras Jalal Mhamad Karim

Abstract: Receiver operating characteristics (ROC) curve analysis is accepted as the most commonly used statistical analysis technique in the medical field to determine a cut-off value for a clinical test and is useful for organizing classiers. In this paper, an ROC Curve is proposed to classify -thalassaemia diseases in Erbil City, Iraq as the research goal, in the classification of diseases, there are, four possible outcomes. If the diseased patient is positive, the case is classied as positive, that means it is counted as a true positive, but if it is classied as negative, that means it counts as a false negative. Should the diseased patient be negative and it is classied as negative, it means that it is country as a true negative, if it is classied as positive, it counts as a false positive. This analysis is applied specifically to a case study in Erbil in Kurdistan Region of Iraq. Results for statistical analysis show that the test (Red_Blood_Cell_Count) has a greater area under curve than the other diagnostic tests after comparison.

Keywords: ROC Curve Analysis, Classification Tool, Medical Test Distinguish, Likelihood Ratio, Criterion Thresholds in optimality

How to Cite?: Aras Jalal Mhamad Karim, "Statistical Study to Classify ?-Thalassaemia diseases in Erbil City at Thalassaemia Center by Using ROC Curve Analysis", Volume 4 Issue 5, May 2015, International Journal of Science and Research (IJSR), Pages: 2955-2961, https://www.ijsr.net/getabstract.php?paperid=SUB154824, DOI: https://dx.doi.org/10.21275/SUB154824