Abstract
Background. Study of morphometric and color characteristics of acid-resistant mycobacteria as objects in digital images of sputum stained by the method of Ziehl-Nielsen.
Materials and methods. Used data on 6,708 acid-resistant mycobacteria isolated from digital images of microscopic preparations of sputum stained by the Ziehl-Nielsen method were used. The analysis of objects was made on 97 color and morphometric signs.
Result. The least important parameters of acid-resistant mycobacteria, as objects in digital images, for the parameterization of the minimum value of R, the maximum value of G, the minimum value of B, the maximum value of S and the maximum value of V, and the most informative parameters in terms of the parameterization of objects are the maximum value of B, the average value of V, the minimum value of H, the minimum value of V and the maximum value of Y and the average value of H. The paper also presents methods for determining the morphometric characteristics of acid-resistant mycobacteria when changing the resolution of digital images, which search for such objects, and change the magnification in which the microscopy of colored sputum samples.
Conclusion. The use of the morphometric and color characteristics obtained as a result of the study will allow the development of algorithms for segmentation of digital microscopic images of sputum stained by the tsilya-Nielsen method, as well as mathematical models for recognition of objects in these images.
Citation
ID:
155801
Ref Key:
narkevich2018vmorphometric