Prediction Analysis of Esophageal Variceal Degrees using Data Mining: Is Validated in Clinical Medicine?
The objective of this study is to assess the feasibility of a data mining association analysis technique in early prediction of esophageal varices in cirrhotic patients and prediction of risky groups candidates for urgent interventional procedure. A manuscript titled “Detection of Risky Esophageal varices using 2D U/S: when to perform Endoscopy”, published in The American Journal of The Medical Science on 21Th of December 2012, to our knowledge it was the first prospective study to assess the degree of esophageal varices by 2D ultrasound using the data mining statistical computed analysis in 673 patients. A descriptive model was generated using a decision tree algorithm (Rapid Miner, version 4.6, Berlin, Germany), the over all accuracy was 95%. Following another 59 patients using statistical analysis to determine the association between esophageal variceal degrees detected by Ultrasound in comparable to Upper Endoscopy, was done. Categorical data were compared using the x2 test, where as continuous variables were compared using Student’s t test. The comparative results accuracy of both two studies was 97.9%.
