The Effect of Surface Defects in Early Caries Assessment

Methods: 45 Extracted human teeth were placed in a partially saturated buffer to form simulated enamel caries lesions. Images of each tooth were captured with both QLF and MDP. The measures ΔF and ΔQ from the QLF methodology, and ΔR and ΔX, from the MDP approach were correlated with average PLM values of lesion depth. Pearson correlations were computed using data from the 27 lesions which did not have surface loss, and then separately based upon the 18 lesions which did display surface loss. MDP variables ΔR and ΔX measure reflected light, whereas QLF variables ΔF and ΔQ measure fluorescence.

Results: When lesions without surface loss were considered, a strong correlation was identified between average PLM (lesion depth) and ÄF (r=-0.765, p < 0.0001), and ΔQ (r=-0.827, p < 0.0001) while a weak but suggestive, although non-significant, correlation was identified between average PLM and ΔR (r=0.369, p = 0.059) and ΔX (r=0.595, p = 0.0011). However, the corresponding correlation was lower, and not statistically significant, when lesions with surface damage were considered for both QLF and MDP measures.

Conclusions: The two QLF-associated measures were found to be strongly correlated with the gold standard average PLM lesion depths when there was no surface loss, but no such relationship was identified among lesions with surface defects.

While a similar pattern was observed when the two MDP-associated measures, ΔR and ΔX, were considered, the level of correlation with the average PLM measure could not be said to differ significantly when lesions with and without surface defects were compared.

NIH DE017875