Fat quantification in the sacroiliac joint syndesmosis: a new semi-automatic volumetric approach

Abstract

Fat is physiologically embedded within the interosseous ligaments in the posterior part of the sacroiliac joint (PSIJ). This composite of fat and ligaments is hypothesized to serve a shock-absorbing, stabilizing function for the sacroiliac joint and the lumbopelvic transition region. Using a novel Python-based software (VolSEQ), total PSIJ volume and fat volume were computed semi-automatically. Differences within the cohort and the viability of the program for the quantification of fat in routine computed tomography (CT) scans were assessed. In 37 CT scans of heathy individuals, the PSIJ were first manually segmented as a region of interest in OSIRIX. Within VolSEQ, ‘fat’ Hounsfield units (− 150 to − 50 HU) are selected and the DICOM file of the patient scan and associated region of interest file from OSIRIX were imported and the pixel sub volumes were then automatically computed. Volume comparisons were made between sexes, sides and ages (≤ 30, 31–64 and > 65 years). PSIJ volumes in both software (VolSeq vs. OSIRIX) were non-different (both 9.7 ± 2.8cm3; p = 0.9). Total PSIJ volume (p = 0.3) and fat volume (p = 0.7) between sexes were non-different. A significant difference in total PSIJ volume between sexes (p < 0.01) but not in fat volume (p = 0.3) was found only in the ≥ 65 years cohort. Fat volume within the PSIJ remains unchanged throughout life. PSIJ volume is sex-dependent after 65 years. VolSEQ is a viable and user-friendly method for sub-volume quantification of tissues in CT.