Reproduzierbarkeit und Validität des ICDAS zur Diagnose der okklusalen Karies - ein Vergleich zwischen der Untersuchung extrahierter Zähne und digitalen Abbildungen

Die Diagnose der Okklusalkaries an digitalen Abbildungen statt an extrahierten Zähnen durchzuführen, könnte den logistischen Aufwand bei Multi-Center-Studien senken und Schulungen sowie die zahnmedizinische Lehre erleichtern. Das visuell-taktile Kariesklassifikationssystem ICDAS ordnet Läsionen unte...

Full description

Saved in:
Bibliographic Details
Main Author: Behrens, Carolin Alexandra
Contributors: Jablonski-Momeni, Anahita (Prof. Dr.) (Thesis advisor)
Format: Doctoral Thesis
Published: Philipps-Universität Marburg 2021
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!

Detection of occlusal caries on digital images instead of extracted teeth could lead to less logistic trouble when organizing multi center studies and could also be helpful in coaching and studies. The visual-tactile International Caries Detection and Assessment System (ICDAS) relates lesions in differing stages with their corresponding extent in hard tooth tissue by interpreting visual signs. Caries is classified from initial to extensive lesions. The present study’s aim is to evaluate whether the visual signs of ICDAS are equally good to detect and interpret on digital images as on extracted teeth. The hypothesis is that diagnostic accuracy of ICDAS detecting occlusal caries on digital images equals to detecting occlusal caries on extracted teeth, validated by histology (gold standard). Methods: The occlusal surfaces of 100 extracted human molars and premolars were scored twice by two observers using ICDAS criteria. Afterwards the teeth were photographed digitally and cut in sections for histologic reasons. Four examiners of differing professional experience scored the digital images on a screen twice at intervals of four weeks using ICDAS criteria. Digital images of the histologic sections were evaluated by all four observers taking a consensus decision using criteria of Downer (gold standard). Intra- and inter-examiner reproducibility was proven by weighted kappa. Correlation of both techniques (extracted teeth/ digital images) with histology was calculated by Spermans correlation coefficient. Sensitivity and specificity were illustrated as ROC curves, Areas under the Curve (AUC) got compared. Results: Intra-examiner reproducibility of ICDAS on extracted teeth varied between ??0.73 - ??0.84 (substantial to almost perfect agreement) and for digital images between ??0.58 - ??0.74 (moderate to substantial agreement). Inter-examiner reproducibility was substantial (??0.62) for extracted teeth and substantial for digital images (?? 0.41 - ??0.60). Correlation with histology was moderate for both methods (extracted teeth: ??0.56 - ??0.60; digital images: ??0.50 - ??0.60). The AUC values of ROC curves of ICDAS on extracted teeth were 0.73 averaged (D1 level) and 0.85 averaged (D3 level). The AUC values of ROC curves of ICDAS on digital images were 0.76 averaged (D1 level) und 0.80 averaged (D3 level). Conclusion: ICDAS applied on digital images to assess occlusal caries does not differ statistically relevant in its validity from ICDAS applied on extracted teeth. Digital images can minimize the logistic challenges of multi center studies, contribute a larger number of examiners and simplify training and education.