Differential Count of WBC in Bone Marrow with Novel Features for Disease Diagnosis

Abstract: White Blood Cell (WBC) differential counts means the different types of cells are measured with various methods. it is necessary in the bone marrow different types of wbc counts in quantity, called differential counts, which provides invaluable information to consultants for disease diagnosis. As the differential wbc counting process is laborious work, an automatic system is preferable. In this paper, we work on investigation whether information about the nucleus alone is sufficient to classify wbcs. This is required because segmentation of nucleus is much easier than the entire cell segmentation, especially in the bone marrow where the wbc density is to much high. In the experiments, a set of manually segmented images of the nucleus are used to decouple segmentation errors. We analyze a set of wbc nucleus based features using Radon-Wavelet transform decomposition. Fivefold cross validation is used in the experiments in which artificial neural networks are applied as classifiers. The classification performances are evaluated by two evaluation measures class wise classification rates. Furthermore, we compare our results with other classifiers and previously proposed nucleus based features. The results show that the features using nucleus alone can be utilized to achieve a classification rate of 81.36% on the test sets. Moreover, the classification is applied to wbc differential count which is used for disease diagnosis.

Keywords: WBC classification,Wavelet Decomposition, wbc differential counts, disease diagnosis.