Multiresolution texture analysis of histopathologic images using ecological diversity measures.

Texture descriptors have been quite popular in medical image analysis, particularly in histopathologic images (HIs), due to the variability of the texture found in such images and the tissue appearance due to irregularity in the staining process. Such variability may exist depending on differences in the staining protocol, such as fixation, inconsistency in the staining condition, and reagents, either between laboratories or in the same laboratory. However, extracting texture features for quantifying HI information in a discriminant way is challenging, given that the distribution of intrinsic properties of such images forms a non-deterministic complex system. This paper proposes a novel method that quantifies such intrinsic properties using ecological diversity measures and discrete wavelet transform. The experimental results on two HI datasets have shown that the proposed method overcomes state-of-the-art shallow and deep methods • Information-theoretical measure of ecological diversity for texture characterization. • Exploitation of independent wavelet sub-band coefficients non-linear interactions. • Unexplored fusion of wavelet features and statistical properties of taxonomic indices. • Texture characterization across HIs with promising results for real-world datasets. [ABSTRACT FROM AUTHOR]