Digital indicators for red cell disorder

Digital light microscopic images of red blood cells (RBC) show a distinctiv e histogram. In contrast, the image histograms of red cells from patients o f thalassaemia and iron deficiency anemia (IDA) show closely placed and amb iguous boundary between the object and the background (near white) peaks. T he RBC image contains a centrally illuminated zone that originates from the biconcavity of the erythrocyte surface. The signature contained in the whi te peak of the histogram is likely to be associated with the characteristic light transmission through this zone. A learning model based on critical l inear separability is presented. The model shows that when judged from bico ncavity alone, thalassaemic cells are different from both normal and LDA ce lls, A related finding was that the fractal dimension of the inner contour of the thalassaemic RBC cells was higher (similar to 1.28) than that of eit her normal (similar to 1.19) or IDA (similar to 1.13) cells.