"Dental Filling Materials"

Read full text
Authors: Aleš Procházka
Cited by: 0

Abstract

"Diffuse reflectance spectroscopy (DRS) is a broadband optical method (400\u20131600 nm) with proven potential for caries detection, yet its ability to identify residual restorative materials \u2014 an unresolved clinical challenge where visual-tactile examination achieves only 60\\% sensitivity and no existing method provides reliable real-time guidance \u2014 has not been systematically investigated. This in vitro study is the first to evaluate DRS as a tool for automated discrimination of dental restorative materials from sound tooth structure. Fifty extracted human premolars and molars together with standardized specimens of four materials (composite resin, universal composite, resin-modified glass ionomer cement, and dental sealant) were measured using a prototype fiber-optic DRS probe, yielding 1080 spectral records across six classes (four different materials, dentin, enamel). A two-layer neural network classifier achieved 98.1\\% accuracy (cross-validation error 0.046), outperforming support vector machine (87.0\\%), Bayesian (88.0\\%), and k-nearest neighbour (90.7\\%) models. The results demonstrate that each material produces a distinct spectral fingerprint clearly separable from healthy hard dental tissues. In clinical practice, a DRS probe could provide objective, operator-independent feedback in real time \u2014 enabling precise removal of old restorations without sacrificing sound dentin and targeted localization of obturation remnants during endodontic retreatment. Future work targets in vivo validation under intraoral conditions, probe miniaturization for endodontic access, and construction of a comprehensive spectral reference database for routine clinical deployment."

Fields of Study & Subjects

Dental materials and restorations Spectroscopy Techniques in Biomedical and Chemical Research Optical Imaging and Spectroscopy Techniques

Publication details

Journal / source
IEEE DataPort
Publication date
15 March 2026
Language
EN
Document type
Dataset