ARTIFICIAL INTELLIGENCE MODELS MAY MATCH THE ACCURACY OF MANUAL SEGMENTATION IN CBCT MEASUREMENTS

Area of Science:

• Dentistry
• Radiology
• Artificial Intelligence

Background:

• Cone-beam computed tomography (CBCT) is crucial for dental diagnostics.
• Accurate tooth segmentation in CBCT images is vital for treatment planning.
• Current segmentation methods face challenges in accuracy and efficiency.

Purpose of the Study:

• To systematically review and meta-analyze the accuracy and time efficiency of artificial intelligence (AI) algorithms for tooth segmentation in CBCT images.
• To evaluate the performance of AI-driven tooth segmentation compared to manual methods.

Main Methods:

• Systematic literature search for studies evaluating AI in CBCT tooth segmentation.
• Meta-analysis of reported accuracy metrics (e.g., Dice score, IoU) and time efficiency data.
• Inclusion of studies published up to the review's completion date.

Main Results:

• AI algorithms demonstrate superior accuracy in tooth segmentation across various CBCT datasets.
• AI significantly reduces the time required for tooth segmentation compared to manual delineation.
• Heterogeneity was observed across studies, necessitating further standardization.

Conclusions:

• AI-powered tooth segmentation in CBCT offers a promising advancement for dental diagnostics.
• AI tools can enhance workflow efficiency and diagnostic precision in dentistry.
• Further research is recommended to optimize AI models and validate their clinical utility.