Fourier transform infrared spectroscopy microscopic imaging classification based on multifractal methods.

Multifractal analysis (MFA) based on generalized concepts of fractals has been applied to biological tissues composed of complex structures. In this paper, a new MFA methodology based on the neighborhood spatial correlation (NSC) is proposed for an extracting texture feature. NSC is used to extract spatial features, and the obtained spatial features are combined with spectral features of characteristic absorption peaks (CAPs) to promote more feature information. This spatial-spectral structure is used as a feature to differentiate cholesterol from Fourier transform infrared spectroscopy microscopic imaging of a rabbit artery by a support vector machine classifier. The dataset was collected between 4000 and 720 cm(-1) on rabbit arteries as research objects. The experimental results show that the accuracy of the proposed spatial-spectral structure is higher than that of other multivariate analysis methods (PCA and 2DPCA). The NSC method, compared to the bottom interface method, new bottom interface method, variance method multi-weight method, and neighborhood spatial correlation method, could effectively reduce the influence of speckle noise, and the convergence rate of the weight factor q is not increased. (C) 2017 Optical Society of America