Title: AUTOMATED DIAGNOSIS OF THALASSEMIA BASED ON DATAMINING CLASSIFIERS

Year of Publication: Jun - 2012
Page Numbers: 440-445
Authors: Eyad H. Elshami, Alaa M. Alhalees
Conference Name: The International Conference on Informatics and Applications (ICIA2012)
- Malaysia

Abstract:


Thalassemia is a genetic disease that is commonly found in many parts of the world. It leads to death in most of its major cases so we must control it by determining the persons who trait the thalassemia genes. Complete Blood Control (CBC) is the first and the simplest test which can narrow to the existence of thalassemia. This paper presents an investigation for thalassemia existence by using data mining classifiers depending on CBC. Three data mining classifiers were used in this investigation. Each of the classifiers used to differentiate between thalassemia traits patients- with its different levels-: iron deficiency patients, normal persons, and the patient who suffer from other blood diseases. The experimental results of this investigation were bright with accuracy exceeding 90% and it showed that the critical point which can be as first indicator for the thalassemia existence is MCV £ 77.65.