STSN-Net: Simultaneous Tooth Segmentation and Numbering Method in Crowded Environments with Deep Learning

Area of Science:

• Dentistry
• Medical Imaging
• Artificial Intelligence

Background:

• Accurate tooth segmentation and numbering are crucial for automated dental diagnosis and treatment planning.
• Existing methods may lack the precision required for complex clinical applications.

Purpose of the Study:

• To develop and evaluate a multitask learning architecture for precise tooth segmentation and numbering in panoramic dental X-ray images.
• To improve the efficiency and accuracy of automated dental diagnostic workflows.

Main Methods:

• A novel multitask learning framework integrating a graph convolution network, a detection subnetwork (DSN), and a region segmentation subnetwork (RSSN).
• Feature fusion between DSN and RSSN to enhance boundary regression accuracy.
• Utilized panoramic X-ray images for training and validation.

Main Results:

• The proposed framework achieved high performance across multiple evaluation metrics.
• Top F1 score of 0.9849, Dice metric score of 0.9629, and mean Average Precision (mAP) of 0.9810 (IOU = 0.5).
• Demonstrated significant improvements in tooth segmentation and numbering accuracy.

Conclusions:

• The developed multitask learning framework offers a robust solution for automated tooth segmentation and numbering.
• This technology has the potential to substantially enhance clinical efficiency for dentists.
• The framework shows promise for advancing automated dental diagnosis and treatment planning.