Pre-trained VGG16 model for forensic dental age estimation
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
The practical employment of Machine Learning in Forensic Odontology remains underexplored, especially in the field of age estimation. Age estimation is essential in legal proceedings to protect the rights of individuals without proper documentation, whether for seeking asylum or when caring for a found child. This study aimed to utilize the VGG16 model to read, analyze, and provide classification of tooth development stages of the third molars from a sample of 876 orthopantomograms (OPGs) 10 to 25 years old (447 males and 429 females) from the Portuguese population collected from the ULS Santa Maria, University of Lisbon. The third molars 38 and 48 were used to classify individuals into age groups based on thresholds of 16, 18, and 21 years old. Age estimation was calculated manually using the methods established by Demirjian and by Moorrees, Funning, and Hunt. Furthermore, we trained the VGG16 model to read, analyze, and provide classification of the development stages, and afterwards we evaluated the VGG16 model through overall accuracy, recall, precision, and F-Score. The goal was to compare the accuracy of the traditional age estimation methods established by Demirjian, and by Moorrees, Funning, and Hunt, with a CNN-based approach.
Results
The VGG16 model provided excellent results for cropped images containing only the third molars (38 and 48) and was able to capture the patterns and the features of development stages, so the overall accuracy obtained was greater than 90%. However, to analyze and classify the development stages defined by Demirjian and by Moorrees, Funning, and Hunt, the VGG16 model faced some limitations due to the insufficient sample of OPGs.
Conclusion
The classification of age groups based on the development of third molars 38 and 48 demonstrated promising results with a high degree of accuracy. However, the limited sample size hindered the VGG16 model's ability to accurately differentiate between the various stages of tooth development. To enhance the model's accuracy and reliability, a larger and more diverse dataset is necessary to better capture the nuances of each developmental stage.
