Automatic sex determination of skulls based on a statistical shape model

Area of Science:

• Forensic Anthropology
• Biometric Analysis
• Digital Imaging

Background:

• Skeletal sex determination is crucial in forensic medicine.
• Current methods rely on subjective visual or metric analyses of sexually dimorphic features.
• These traditional methods require expert knowledge and extensive manual measurements.

Purpose of the Study:

• To develop an automated method for sex determination from 3D digital skulls.
• To combine metrical and morphological analysis advantages for improved accuracy.
• To create an easy-to-use tool for sex determination without professional qualifications.

Main Methods:

• Construction of a statistical shape model for 3D digital skulls.
• Projection of high-dimensional skull data into a low-dimensional shape space.
• Application of Fisher discriminant analysis for skull classification within the shape space.

Main Results:

• The automated method achieved high accuracy rates: 95.7% for females and 91.4% for males on a Chinese skull dataset.
• Validation using independent samples confirmed the model's effectiveness.
• Leave-one-out testing demonstrated the method's robust accuracy.

Conclusions:

• The developed automated method provides an accurate and efficient approach to skeletal sex determination.
• This technique simplifies the process, reducing the need for specialized expertise and manual measurements.
• The method shows significant potential for application in forensic anthropology and related fields.