Hyoid bone CT analysis estimates sex with 93.8 percent accuracy in forensic study

A retrospective study published in the International Journal of Legal Medicine reports that computed tomography analysis of the hyoid bone can classify sex with 93.8 percent accuracy. Researchers in Turkey analyzed cervical vertebrae CT scans from 660 randomly selected cases, evenly split between 330 males and 330 females, aged 15 to 80 years and collected between 2021 and 2024, applying specific exclusion criteria. Because age and sex estimation are crucial for identification in forensic investigations of skeletal remains, the team set out to evaluate anthropometric measurements, fusion degrees, and radiodensity of the hyoid bone in this large sample.

Using 3D Slicer software, the researchers reconstructed the imaging data into three-dimensional models to perform osteometric measurements, stage the degree of fusion of the bone's segments, and quantify both volume and radiodensity. The statistical toolkit included independent t-tests to assess sexual dimorphism, intra-class correlation coefficients to evaluate observer reliability, binary logistic regression for sex estimation, and stepwise multiple linear regression to derive age-estimation models. The results showed pronounced anatomical differences between the sexes.

Morphometric and volume measurements were significantly higher in males. Radiodensity told a different story: females had lower radiodensity values across all ages, a gap that was particularly evident in the older groups. The study was conducted on a Turkish population sample, so the findings are population-specific within the scope analyzed.

For sex estimation, the multivariate stepwise binary logistic regression model correctly classified sex with 93.8 percent accuracy. Observer reliability, assessed through intra-class correlation coefficients, supported the consistency of the measurement protocol, an important consideration for any method intended for forensic casework where different examiners must reach comparable conclusions from the same scans. Age estimation proved more limited.

The multivariate stepwise linear regression models explained 51.2 percent of the variance in females and 45.4 percent in males, meaning that roughly half of the individual variation in the measured age-related traits remained unaccounted for. The authors are explicit about the practical implication: because the age-estimation models have limited explanatory power, an estimated age range is more reliable than an exact age when the hyoid bone is used. The study's contribution lies in its multiparametric design.

Rather than relying on a single metric, it combined size measurements, fusion staging, volume, and radiodensity from the same three-dimensional reconstructions, capturing complementary dimensions of skeletal change from a single, non-destructive imaging dataset. The authors describe CT-based multidimensional analysis of the hyoid bone as a robust method for sex estimation and a supportive tool for age estimation, useful for corroborating conclusions drawn from other indicators rather than standing alone as an age source. In identification contexts, where narrowing the pool of potential matches is a first step, a bone that reliably indicates sex while providing supportive age information can still materially advance an investigation.

As with all population-based forensic models, the approach was developed from a specific sample and time window, and the reported accuracies apply within that retrospective study design. Within its stated scope, the study demonstrates that the hyoid bone exhibits pronounced sexual dimorphism in a Turkish population and that routine CT scans can be repurposed to extract forensic information from this structure through three-dimensional anthropometry and radiodensity analysis.

Subject of Research: Computed tomography-based forensic age and sex estimation using hyoid bone anthropometry and radiodensity

Article Title: Computed tomography-based age and sex estimation through hyoid bone anthropometry and radiodensity analysis

Article References: Çelik, D. G., Balandız, H., Bozlar, U., Özsoy, S., & Kaya, B. (2026). Computed tomography-based age and sex estimation through hyoid bone anthropometry and radiodensity analysis. International Journal of Legal Medicine. https://doi.org/10.1007/s00414-026-04054-y

Image Credits: AI Generated

DOI: 10.1007/s00414-026-04054-y

Keywords: forensic anthropology, hyoid bone, computed tomography, sex estimation, age estimation, radiodensity, sexual dimorphism, 3D Slicer, logistic regression, Turkish population, Computed, tomography-based