Thanks to technical progress and the availability of virtual data, sex estimation methods as part of a biological profile are undergoing an inevitable evolution. Further reductions in subjectivity, but potentially also in measurement errors, can be brought by approaches that automate the extraction of variables. Such automatization also significantly accelerates and facilitates the specialist’s work. The aim of this study is (1) to apply a previously proposed algorithm (Kuchař et al. 2021) to automatically extract 10 variables used for the DSP2 sex estimation method, and (2) to test the robustness of the new automatic approach in a current heterogeneous population. For the first aim, we used a sample of 240 3D scans of pelvic bones from the same individuals, which were measured manually for the DSP database. For the second aim a sample of 108 pelvic bones from the New Mexico Decedent Image Database was used. The results showed high agreement between automatic and manual measurements with rTEM below 5% for all dimensions except two. The accuracy of final sex estimates based on all 10 variables was excellent (error rate 0.3%). However, we observed a higher number of undetermined individuals in the Portuguese sample (25% of males) and the New Mexican sample (36.5% of females). In conclusion, the procedure for automatic dimension extraction was successfully applied both to a different type of data and to a heterogeneous population.
5 June 2026 in International journal of legal medicine
Background: Biological sex estimation is a cornerstone of the forensic biological profile, and the adult pelvis remains the most informative skeletal region due to pronounced sexual dimorphism. At the same time, three-dimensional documentation of skeletal evidence is rapidly expanding beyond computed tomography and laboratory scanners toward inexpensive, accessible approaches such as photogrammetry—with mobile phone cameras. While the reliability of photogrammetry for osteological documentation has been supported, most established pelvic sexing approaches still assume measurements in absolute units or rely on expert scoring of morphological traits. Methods: We propose and validate a scale-free workflow for sex estimation from unscaled 3D pelvic models. The final approach is based on within-specimen log-ratios of seven collected pelvic measurements rather than on absolute units. Model development used a pooled reference dataset of 2,321 known-sex os coxae from documented dry-bone and CT-derived sources, and external implementation testing targeted unscaled photogrammetric pelvic models acquired under realistic conditions ( n = 103). Results: In the Coimbra photogrammetric subset, the final scale-free model achieved 97.1% binary accuracy, 97.1% balanced accuracy, and an area under curve of 0.997. Under conservative strict thresholds, 90.3% of cases received a definitive classification, all of which were correct, while the remaining 9.7% were retained as indeterminate. Conclusions: Scale-free pelvic sex estimation supports the forensic use of unscaled photogrammetric 3D bone models in situations where reliable external calibration is unavailable. The method does not replace standard calibrated osteometry, but it provides a practical and defensible alternative for low-cost digital documentation, remote analysis, and triage-oriented forensic workflows.
7 October 2026 in Journal of The Royal Society Interface
Walking, posture and childbirth all depend on the pelvic ring, yet human pelvic dimorphism is usually interpreted from static birth-canal dimensions. We asked whether sex differences in pelvic function are explained by overall size and static pelvimetry or by available mechanical deformation pathways. We analysed 278 anatomically standardized lumbopelvic finite-element models built from clinical computed tomography scans. In each model, modal decomposition separated pelvic deformation into a shared, ranked set of pathways, allowing us to compare pathway stiffness, load recruitment and sensitivity to ligamentous and cartilaginous constraints. Overall pelvic size accounted for much of the baseline stiffness spectrum, confirming a strong size-dependent component. However, size- and age-adjusted sex differences remained: male pelves showed approximately 20% higher eigenvalues than female pelves in modes 2 and 9. Loading probes showed a common backbone in both sexes during habitual support, whereas inlet opening revealed a different response organization, with female pelves distributing deformation across a broader, coordinated set and a higher effective number of participating modes. The principal inlet-opening pathway was most sensitive to the pubic symphysis and anterior sacroiliac ligamentous constraints. Pelvic dimorphism, therefore, lies not only in canal shape, but in the organization of mechanical pathways that balance support with deformation.
Objective: Standard osteoporosis assessment using dual-energy X-ray absorptiometry (DXA) provides limited 2D measures (BMD, T-score), potentially missing localized bone density differences relevant to fracture risk. Opportunistic CT allows comprehensive 3D volumetric bone density analysis. This study presents a detailed 3D statistical model of CT-derived apparent volumetric bone mineral density (CT-vBMD) covering the L4 and L5 vertebrae, pelvic bones, sacrum, and proximal femur and examines spatial distributions and associations with age and sex. Materials and methods: Retrospective analysis included 330 routine abdominal CT scans (ages 20–91; approximately balanced by sex). Bones were segmented with a semi-automated workflow, internally calibrated, registered to bone-specific population templates, and projected onto corresponding tetrahedral meshes. Whole-bone averages were compared across age groups and between sexes. For each element, we calculated mean vBMD, variability, correlations with age and sex, and cluster-defined low-density and age-sensitive subregions. Results: The global spatial pattern was similar in females and males. The highest CT-vBMD values were localized mainly to cortical load-bearing zones, whereas the lowest values were concentrated in trabecular spongiosa. Older age was associated with lower whole-bone CT-vBMD in all bones and with stronger negative associations in trabecular-rich regions, such as the subchondral acetabulum, femoral neck, sacral ala, and vertebral bodies. Cortical-dominant load-bearing regions showed weaker negative associations with age. Conclusion: This framework maps anatomically localized associations between age and CT-vBMD beyond the regions routinely assessed by DXA. Opportunistic CT mapping may complement standard osteoporosis assessment by providing spatial information for the pelvis, sacrum, lower lumbar spine, and proximal femur.
Bone radiodensity is often reduced to global Hounsfield Unit (HU) summaries, although its spatial distribution may contain biologically relevant information. This study introduced a template-based framework for spatial radiodensity mapping of the adult human hyoid and tested whether spatial descriptors capture age- and sex-related variation better than whole-bone HU measures. Clinical head-and-neck CT scans from 300 adults, including 180 females and 120 males, were analyzed. Hyoids were segmented, normalized into a common anatomical template, and divided into 1500 supervoxels for spatial analysis. Scalar HU summaries were compared with PCA-based spatial descriptors, and regional effects were assessed using supervoxel-wise models with false-discovery-rate correction. The registered radiodensity maps showed spatial coherence, with neighboring supervoxels displaying more similar profiles than distant ones. In this cohort, spatial descriptors outperformed mean HU for age prediction, although performance remained limited (R2 = 0.255). Sex classification was moderate (ROC AUC = 0.810; balanced accuracy = 0.725). After the spatial features were adjusted for native volume, sex-classification AUC fell to 0.671, and none of the 51 unadjusted regional sex associations survived volume adjustment. Age-related associations involved the hyoid body, the body–greater horn transition, and the proximal-to-mid greater horns. These findings demonstrate spatial coherence in the registered maps but show that the sex-related pattern is strongly size-dependent.