'CT and CT image-based texture image analysis in radiological diagnostics of allergic fungal rhinosinusitis'

Background: This prospective study is focused on evaluating radiological properties of AFRS. We analysed specific CT features related to the presence of AFRS, as well as explored the possible usefulness of the texture image analysis (TIA) as an additional diagnostical parameter.

Methods: The CT images of maxillary sinuses of 37 adult patients diagnosed with chronic rhinosinusitis were analysed for homogeneity, high-attenuation areas, density of the soft tissue mass, bony wall thickness and density. TIA included assessment of uniformity, contrast, homogeneity and entropy of sinus content.

Results: In the F+ group, soft tissue mass was significantly more non-homogeneous, high-attenuation areas were more prevalent, while soft tissue densities were higher. The sinus wall showed a tendency towards decreased thickness and significantly higher density in the F+ group. Among TIA parameters only homogeneity was significantly lower in the F+ group.

Conclusions: Presence of fungi should be suspected when the sinus is filled with a non-homogenous soft tissue content of a high CT density not necessarily presented as clearly visible hyperattenuation material. Additional criteria in radiological diagnostics of AFRS should encompass assessment of sinus bony wall density. TIA may serve as a tool for quantitative assessment of subjective CT features such as homogeneity of the soft tissue mass for investigative purposes. However, other TIA parameters showed limited potential.