Accuracy and Time Efficiency of Automated Tooth Segmentation in Dental Imaging-A Systematic Review and Meta-Analysis

Area of Science:

• Dentistry
• Medical Imaging
• Artificial Intelligence

Background:

• Tooth segmentation is crucial for dental diagnostics and treatment planning.
• Manual segmentation is time-consuming and prone to inter-observer variability.
• Automated methods using artificial intelligence (AI) offer potential for improved efficiency and accuracy.

Purpose of the Study:

• To systematically review and meta-analyze the accuracy and efficiency of AI-based automated tooth segmentation methods.
• To compare AI methods against manual segmentation and ground truth techniques.
• To assess the performance of different AI algorithms in tooth segmentation.

Main Methods:

• A comprehensive systematic search was conducted across multiple databases (MEDLINE, Cochrane, ScienceDirect, etc.) up to January 2024.
• The Quality Assessment Tool for Diagnostic Accuracy Studies-2 (QUADAS-2) was used for risk of bias assessment.
• Meta-analysis included 37 studies, evaluating sensitivity, specificity, Dice coefficient, and Hausdorff distance.

Main Results:

• AI and manual methods showed comparable Dice segmentation coefficients (SMD = 0.05, p = 0.9) and Hausdorff distances.
• Ground truth AI algorithms significantly outperformed proposed AI algorithms in Dice coefficient (SMD = 2.42, p < 0.00001).
• AI algorithms demonstrated significantly greater speed compared to manual methods.

Conclusions:

• AI-based tooth segmentation performs comparably to manual segmentation in terms of accuracy.
• Recent AI algorithms show superior performance over ground truth algorithms.
• The certainty of evidence is very low due to bias and heterogeneity, necessitating further high-quality research.