Synchrotron and AI-enabled large-scale 3D morphometric analysis of osteocyte lacunae reveals disease-associated microstructural remodeling in human trabecular bone

Bone fragility in Osteoporosis and following COVID-19 infection has been increasingly reported at the tissue level; however, direct human evidence linking these conditions to alterations in osteocyte lacunar microarchitecture remains limited. Osteocyte lacunae form the mechanosensory network of bone, where geometry modulates local strain amplification and fluid flow. Quantitative characterization of disease-associated lacunar remodeling in human trabecular bone has been constrained by imaging resolution and sample size. Trabecular bone from 48 human female femoral heads (Healthy, Osteoporosis, and post–COVID-19; n = 16 per group) was analyzed using synchrotron phase-contrast micro-CT combined with AI-based recognition of relevant features. Approximately 8.96 million lacunae were quantified, enabling large-scale extraction of three-dimensional morphometric descriptors. Osteoporosis was associated with significantly larger lacunae than post–COVID-19 bone (p < 0.05), reflected by increased volume, surface area, and semi-minor axis length, while healthy bone showed intermediate values. Lacunar sphericity and BV/TV were significantly reduced in Osteoporosis (p = 0.019), whereas lacunar density did not differ among groups, indicating geometric remodeling rather than changes in lacunar abundance. These findings establish a quantitative benchmark for disease-associated lacunar remodeling in human trabecular bone and demonstrate that osteocyte lacunar geometry captures microstructural signatures not reflected by densitometric measures alone.