Auto-Segmentation and Quantification of Non-Cavitated Enamel Caries Imaged with Swept-Source Optical Coherence...

Area of Science:

• Biomedical Engineering
• Dental Research
• Optical Imaging

Background:

• Optical Coherence Tomography (OCT) is used in dental research for enamel demineralization assessment.
• Current OCT software limitations restrict clinical diagnostic use to qualitative analysis.
• A need exists for efficient, validated algorithms for quantitative enamel demineralization analysis.

Purpose of the Study:

• To develop and validate a Graphical User Interface (GUI) MATLAB algorithm for processing and quantitative analysis of demineralized enamel using OCT.
• To enable segmentation and measurement of inter- and intra-prismatic demineralization zones.
• To assess the algorithm's accuracy and potential for real-time dental diagnostics.

Main Methods:

• Human enamel samples with artificial demineralization were scanned using OCT.
• Region of Interest (ROI) frames were extracted for analysis.
• An intensity threshold colormap segmented inter- (Ie) and intra- (Ia) prismatic demineralization.
• Quantitative measurements included average demineralized depth, line profile, and integrated reflectivity.
• Algorithm verification used real and simulated OCT frames.

Main Results:

• A strong correlation (R² > 0.97) was observed between automated and Excel measurements for average demineralization depth.
• The algorithm successfully segmented and quantified enamel demineralization zones.
• Verified accuracy of the developed quantitative analysis.

Conclusions:

• OCT image segmentation and quantification of enamel demineralization zones are feasible.
• The developed algorithm provides accurate quantitative assessment of demineralized enamel.
• This technology holds potential for future real-time dental diagnostics with an oral probe OCT.