Statistical classification of the states of biological cells treated with carbon nanotubes based on AFM-images of cell surface

Abstract

The method of statistical classification of biological cells, treated with carbon nanotubes, based on images of cell surface obtained with an atomic force microscope (AFM) is proposed. Each scan line of the original AFM image is considered as a random sequence realization, and the discrete Fourier transform is applied to compute its spectral features. After smoothing, the map of spectral estimates is formed. The informative features are computed as the medians for the set of the spectrogram values. Classification of four classes of cells (control and treated with carbon nanotubes, after 1 hour and 24 hours of incubation) was carried out by the obtained informative features using the decision trees method. The proposed method provides a sufficiently high accuracy classification of cell states after the treatment with carbon nanotubes