Segmentation of bone marrow biopsies by mathematical morphology in color spaces

Abstract

Anatomic and pathologic reports of histological slides give results indicating the presence of trabeculae, hematopoietic and fat cells. This allows evaluating the grade of some metabolic disorder, comparing normal values with pathological ones. Generally, quantifications are carried out by simple visual inspection presenting big inter-and intra-user differences. This work presents an automatic method for detecting different types of tissues in bone marrow biopsies using Mathematical Morphology in color spaces, in order to determine the marrow cellularity accurately. The method was tested with 30 images of bone marrow biopsies presenting different distributions of cellularity.