Novel Procedure for Automatic Registration between Cone-Beam Computed Tomography and Intraoral Scan Data Supported...

Area of Science:

• Dentistry
• Medical Imaging
• Computer Vision

Background:

• Intraoral scans and CBCT are standard for 3D dental models but have limitations.
• Integrating these datasets offers potential for improved diagnostic accuracy and treatment planning.

Purpose of the Study:

• To develop and validate a novel 3D registration approach for harmonizing intraoral scan and CBCT data.
• To assess the accuracy and clinical potential of the proposed data fusion technique.

Main Methods:

• Pre-processing involved Mask-RCNN for tooth segmentation from CBCT and chromatic classification for intraoral scan data.
• Dental crown-based alignment was performed using RANSAC and ICP algorithms.
• Accuracy was evaluated using Euclidean distance and expert landmark measurements.

Main Results:

• The proposed method achieved a computed error of 0.234 ± 0.019 mm between intraoral scan and CBCT data, below the 0.3 mm CBCT voxel size.
• Expert evaluation showed average measurement discrepancies ranging from 0.368 to 1.079 mm for landmarks.

Conclusions:

• The novel 3D registration technique effectively integrates intraoral scans and CBCT data with high accuracy.
• This approach shows significant promise for enhancing dental diagnostics and treatment planning by providing a comprehensive 3D perspective.