Alveolar Bone Segmentation Methods in Assessing the Effectiveness of Periodontal Defect Regeneration Through Machine...

Area of Science:

• Dentistry
• Medical Imaging
• Regenerative Medicine

Background:

• Periodontal defects require precise imaging for effective treatment.
• Cone-beam computed tomography (CBCT) is a key imaging modality.
• Digital workflows are increasingly important in periodontal regeneration.

Purpose of the Study:

• To evaluate segmentation methods for CBCT images of alveolar bone.
• To assess the effectiveness of these methods in digital workflows for periodontal defect regeneration.

Main Methods:

• Systematic literature search (May 2024-June 2025) using MeSH terms across major databases.
• Adherence to PRISMA-ScR guidelines and QUADAS-2 for bias assessment.
• Included 23 studies, focusing on segmentation techniques for alveolar bone defects.

Main Results:

• Deep learning methods, particularly U-Net, were most common.
• Segmentation performance, measured by Dice Similarity (DC) index, ranged from 76% to 98%.
• Significant differences noted in segmenting healthy vs. defective alveolar bone.

Conclusions:

• Accurate segmentation of periodontal defects is critical for scaffold design.
• Enhanced deep learning methods are needed for improved segmentation accuracy.
• This forms a crucial first step in patient-specific scaffold workflows.