Algorithm to assess cranial suture fusion with varying and discontinuous mineral density

One of the most difficult challenges in medical imaging is the accurate segmentation of mineralized tissues. This process is complicated when studying developmental or regenerative processes due to the changes in mineral density that these tissues undergo over time. To address these limitations an algorithm was developed to enable the use of computed tomography (CT) to study tissues of varying and heterogeneous mineralization. To examine and validate this algorithm a study was performed on the development of murine cranial sutures. C57Bl/6J mice ranging in age from 6 to 25 days were imaged using micro-CT (μCT). The algorithm was developed to segment the bones of both the posterior frontal (PF) and coronal (COR) sutures. For the curved COR suture, an addition to the algorithm was developed to reconstruct images that were perpendicular to the suture about all three axes. The algorithm showed excellent linear correlation (R (2) > 0.96) with serial histomorphometry and nearly a 1:1 relationship between all measures. The algorithm was validated with serial histology. The algorithm showed that the PF suture fused between days 12 and 20 but then showed a significant increase in bone volume after day 20. The algorithm developed provides an accurate method to segment the irregular sutures of the mouse calvaria.