Information fusion for infant age estimation from deciduous teeth using machine learning.

Area of Science:

• Forensic Anthropology
• Pediatric Dentistry
• Machine Learning

Background:

• Dental development is a reliable marker for infant age estimation.
• Traditional methods struggle to integrate diverse dental features effectively.
• Machine learning (ML) offers a potential solution for enhanced accuracy.

Purpose of the Study:

• To develop and validate an ML model for infant age estimation using deciduous teeth.
• To assess the informativeness of various dental features (length, mineralization, alveolar stage).
• To evaluate the benefit of combining information from multiple teeth and features.

Main Methods:

• Utilized a dataset of 114 infant skeletons (5 months gestation to 3 years).
• Employed a Multilayer Perceptron (MLP) model, a type of artificial neural network.
• Validated the model using a leave-one-out cross-validation protocol.
• Experimentally analyzed individual and combined dental features.

Main Results:

• Fusion of dental variables yielded more accurate age estimates (RMSE = 66 days) than individual variables (RMSE = 101 days).
• Incorporating multiple teeth significantly reduced the root mean square error (RMSE) compared to using a single tooth.
• The ML approach demonstrated superior accuracy and robustness in age estimation.

Conclusions:

• ML-based methods offer significant advantages for infant age estimation.
• Integrating multiple dental features and teeth enhances accuracy and reliability.
• This study highlights the potential of ML in forensic and pediatric age assessment.