Fractal analysis tools for early assessment of liver inflammation induced by chronic consumption of linseed, palm and sunflower oils

Abstract

Objective: Inflammation is a biological response of tissue to harmful stimuli. A high-fat diet was linked to low-grade chronic liver inflammation, which can further lead to more severe health conditions. It is crucial to assess the intensity of inflammation and structural tissue changes to reduce the subjective judgment by the examiner. We propose fractal-based methods for early-stage low-degree liver inflammation grading. Methods: We have randomly divided 40 C57BL/6 female mice into four groups (control, linseed oil, palm oil, sunflower oil). After 100 days, animals were euthanized, and liver tissue collected for analyses. We performed calculations of fractal dimension, fractal lacunarity, multifractal spectra, local fractal dimension, and particle metrics, applicable to tissue segmentation and grading. Results: Pathohistological analysis of some liver tissue showed a low-grade inflammatory infiltrate around the portal vein of experimental groups subjected to different high-fat diets. Differences in fractal dimension and lacunarity of the inflamed tissue were, in most cases, statistically significant between the high-fat diet groups. Both the observed intensity and area of inflammation were lowest for the sunflower oil. The results of standard fractal analysis, local fractal analysis, and particle analysis were in an excellent agreement. Conclusions: This study demonstrated the efficiency of the fractal analysis based tools in the quantification of complexity and early-stage structural changes in inflamed liver tissue, which could potentially be used in the diagnostic workup of inflammation in the liver. The presented methods could be implemented within a wider scope computer-aided diagnostics system in a very straightforward manner.