A Digital Shade-Matching Device for Dental Color Determination Using the Support Vector Machine Algorithm

Area of Science:

• Biomedical Engineering
• Computer Science
• Materials Science

Background:

• Accurate dental shade matching is crucial for aesthetic restorations.
• Traditional shade-matching methods using physical tabs are subjective and prone to error.
• Digital solutions are needed for objective and reproducible color determination in dentistry.

Purpose of the Study:

• To develop and validate a digital shade-matching device for precise dental color determination.
• To employ the support vector machine (SVM) algorithm for accurate classification of tooth shades.
• To evaluate the performance and accuracy of the developed device compared to existing methods.

Main Methods:

• Modified intraoral camera with a cross-polarization scheme to minimize external light interference.
• Utilized a precise robot arm and a specialized jig for consistent positioning of VITA 3D-master (3D) shade tabs.
• Implemented color calibration using standard colors and developed a database with five devices for SVM training.
• Tested shade matching accuracy by measuring 3D shade tabs with three additional devices.

Main Results:

• Achieved an average matching accuracy exceeding 90% across all tested devices.
• Recorded a failure rate of less than 1% over 10 measurements per device.
• Demonstrated superior classification performance of the SVM algorithm compared to logistic regression, random forest, and k-nearest neighbors via cross-validation.

Conclusions:

• The developed digital shade-matching device, powered by the SVM algorithm, provides a highly accurate and reliable method for quantitative tooth color measurement.
• The proposed system offers an optimal solution for objective dental shade analysis, improving aesthetic outcomes.
• The cross-polarization technique and precise hardware components contribute to the robustness and accuracy of the digital shade-matching system.