Sign in Subscribe
Artificial intelligence

Machine Learning Techniques for Sex Estimation in Forensic Examination

Strategic Insights

— 1 min read

Researchers at the Council of Forensic Medicine have conducted a study to determine the effectiveness of various machine learning techniques for sex estimation in forensic examination. The study used CT images of 485 cases from the Turkish population, with 248 males and 237 females. The results showed that sternum measurements or images can be used for sex estimation in the Turkish population, with the best-performing model being linear discriminant analysis (LDA). The study also compared the usability of deep neural networks (DNN) in sternum images, finding that while DNN had lower performance, there was no statistically significant difference between the area under curves (AUCs) of DNN and LDA.

Key Takeaways:

  • The study aimed to determine the effectiveness of various machine learning techniques for sex estimation in forensic examination, using CT images of 485 cases from the Turkish population.
  • The results showed that sternum measurements or images can be used for sex estimation in the Turkish population, with a statistically significant difference between male and female cases.
  • The best-performing model was linear discriminant analysis (LDA), followed by logistic regression, naive Bayes, XGBoost, random forest, and KNN, respectively.
  • Deep neural networks (DNN) had lower performance, but there was no statistically significant difference between the area under curves (AUCs) of DNN and LDA.
  • The study concluded that sternum measurements or images can be used for sex estimation in the Turkish population, and that machine learning techniques can be effective in forensic examination.

Statistics:

  • 485 cases were used in the study, with 248 males and 237 females.
  • The sagittal plane manubrial length (M) and corpus sternal length (CSL) were measured in 485 cases.
  • The coronal plane manubrial width (MW), sternal body width at first sternebra (CSW1), and sternal body width at third sternebra (CSW3) were measured in 485 cases.
  • The sternal index (SI), corpus sterni length (CSL), and sternal area (SA) were calculated in 485 cases.

Sources:

  • "Sex estimation from the sternum in Turkish population using various machine learning methods and deep neural networks." The Egyptian Journal of Radiology and Nuclear Medicine, 2025,56(1):1-9.
  • "New Machine Learning Research Has Been Reported by Researchers at Council of Forensic Medicine (Sex estimation from the sternum in Turkish population using various machine learning methods and deep neural networks)." Journal of Engineering. October 20, 2025; p 2044.

Read more