Fuzzy Feature Representation for White Blood Cell Differential Counting in Acute Leukemia Diagnosis

A fuzzy feature representation for white blood cell differential counting is proposed to diagnose types of acute leukemia. The accuracy of diagnosis is higher than that by numerical features by dealing with uncertainty of white blood cell features and inflexibility of diagnosing. Experiments on acute leukemia diagnosis use 120 acute leukemia images and fuzzy decision tree method with the accuracy rate of diagnosis is 84% using fuzzy features and 76.6% using numerical features. Given the importance of accurate diagnosis of acute leukemia in patients, this proposal is essential and planned to be introduced in an Indonesian hospital.