Scholarlink

Purpose <p>The estimation of sex is the first and foremost step in developing a reliable biological identity during the examination of skeletal remains. The purpose of the study is to assess the usability of the sacrum for sex determination in a Turkish population through machine learning.</p> Methods <p>We performed a retrospective examination of the CT images of 540 individuals. DICOM images, including the sacrum, were obtained from patients aged 18–79 years. The following parameters were measured: Anterior Sacral Length (ASL), Posterior Sacral Length (PSL), Anterior Sacrococcygeal Length (ASCL), Posterior Sacrococcygeal Length (PSCL), Maximum Anteroposterior Diameter (MAPD), and Maximum Transverse Diameter (MTD). Eight different machine learning algorithms were applied for sex estimation. To enhance reliability, fivefold cross-validation was used.</p> Results <p>Sacral morphometric parameters were statistically significantly higher in males than in females (<i>p</i> &lt; 0.001). Among the algorithms tested, logistic regression achieved the highest classification accuracy for sex at 85%. Based on these results, logistic regression provided the best performance for sex estimation from sacral measurements. Further, the maximum anteroposterior diameter at the sacral base was the most discriminative parameter.</p> Conclusion <p>Based on our analysis shows that the sacrum is a reliable anatomical structure for predicting sex in cases of high sexual dimorphism with an accuracy rate of 85%. The sacrum can serve as a powerful reference for sex estimation in forensic cases and anthropological applications. Furthermore, the resulting morphometric records can be stored as a useful resource for preoperative planning and anatomical evaluation for clinics and regional surgeons.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Sex estimation from sacral anatomy in Turkish adults: a machine learning–based analysis

  • 期刊论文

  • 发表日期：

    2026年3月18日

查看全文
  • Busra Gul Ayturk,
  • Ali Keles,
  • Usame Omer Osmanoglu

摘要

Purpose

The estimation of sex is the first and foremost step in developing a reliable biological identity during the examination of skeletal remains. The purpose of the study is to assess the usability of the sacrum for sex determination in a Turkish population through machine learning.

Methods

We performed a retrospective examination of the CT images of 540 individuals. DICOM images, including the sacrum, were obtained from patients aged 18–79 years. The following parameters were measured: Anterior Sacral Length (ASL), Posterior Sacral Length (PSL), Anterior Sacrococcygeal Length (ASCL), Posterior Sacrococcygeal Length (PSCL), Maximum Anteroposterior Diameter (MAPD), and Maximum Transverse Diameter (MTD). Eight different machine learning algorithms were applied for sex estimation. To enhance reliability, fivefold cross-validation was used.

Results

Sacral morphometric parameters were statistically significantly higher in males than in females (p < 0.001). Among the algorithms tested, logistic regression achieved the highest classification accuracy for sex at 85%. Based on these results, logistic regression provided the best performance for sex estimation from sacral measurements. Further, the maximum anteroposterior diameter at the sacral base was the most discriminative parameter.

Conclusion

Based on our analysis shows that the sacrum is a reliable anatomical structure for predicting sex in cases of high sexual dimorphism with an accuracy rate of 85%. The sacrum can serve as a powerful reference for sex estimation in forensic cases and anthropological applications. Furthermore, the resulting morphometric records can be stored as a useful resource for preoperative planning and anatomical evaluation for clinics and regional surgeons.