A New Strategy for the Morphological and Colorimetric Recognition of Erythrocytes for the Diagnosis of Forms of Anemia based on Microscopic Color Images of Blood Smears

The detection of red blood cells based on morphology and colorimetric appearance is very important in improving hematology diagnostics. There are automatons capable of detecting certain forms, but these have limitations with regard to the formal identification of red blood cells because they consider certain cells to be red blood cells when they are not and vice versa. Other automata have limitations in their operation because they do not cover a sufficient area of the blood smear. In spite of their performance, biologists have very often resorted to the manual analysis of blood smears under an optical microscope for a morphological and colorimetric study. In this paper, we present a new strategy for semi-automatic identification of red blood cells based on their isolation, their automatic color segmentation using Otsu's algorithm and their morphology. The algorithms of our method have been implemented in the programming environment of the scientific software MATLAB resulting in an artificial intelligence application. The application, once launched, allows the biologist to select a region of interest containing the erythrocyte to be characterized, then a set of attributes are computed extracted from this target red blood cell. These attributes include compactness, perimeter, area, morphology, white and red proportions of the erythrocyte, etc. The types of anemia treated in this work concern the iron-deficiency, sickle-cell or falciform, thalassemia, hemolytic, etc. forms. The results obtained are excellent because they highlight different forms of anemia contracted in a patient.

@article{ALICO2020,
  title     = {A New Strategy for the Morphological and Colorimetric Recognition of Erythrocytes for the Diagnosis of Forms of Anemia based on Microscopic Color Images of Blood Smears},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {11},
  number    = {7},
  year      = {2020},
  publisher = {The Science and Information Organization},
  author    = {J. Nango ALICO and Sie OUATTARA and Alain CLEMENT},
  doi       = {10.14569/IJACSA.2020.0110762},
  url       = {https://doi.org/10.14569/IJACSA.2020.0110762}
}