METHODS: Tooth wear status of NPC survivors were clinically assessed using the Exact Tooth Wear Index. A tooth was graded to have severe wear when more than one-third of its buccal/occlusal/lingual surface had dentine loss. At the subject-level, percentages of anterior/posterior/all teeth with severe wear were calculated. Age, number of teeth, flow-rate/buffering capacity/pH of stimulated whole (SWS) and parotid (SPS) saliva's were collected. Correlation and multiple-linear regression tests were performed at the significance level α = 0.05.

RESULT: Sixty-eight participants (mean age of 60.0 ± 8.9), 697 anterior and 686 posterior teeth were examined with a mean of 10-years post-radiotherapy. Severe tooth wear was found in 63 (92.6 percent) participants, 288 anterior and 83 posterior teeth. The mean percentage of anterior/posterior/all teeth with severe wear were 42.3 ± 28.1, 14.5 ± 19.9 and 30.0 ± 21.7. Anterior teeth, particularly the incisal surface of central incisors were most affected. The mean flow-rate of SWS and SPS were 0.1 ± 0.1 ml/min and 0.03 ± 0.07 ml/min respectively. Thirty (44.1 percent) and 48 (70.6 percent) participants were found to have low/no buffering capacity of SWS and SPS respectively. Multiple-regression analyses revealed the SWS flow-rate was associated with the percentage of anterior teeth with severe wear (p=0.03).

CONCLUSION: Anterior tooth wear is a significant dental problem among NPC survivors and was associated with hypo-salivation.

METHODS: Frontal view intraoral photographs fulfilling selection criteria were collected. Along the gingival margin, the gingival conditions of individual sites were labelled as healthy, diseased, or questionable. Photographs were randomly assigned as training or validation datasets. Training datasets were input into a novel artificial intelligence system and its accuracy in detection of gingivitis including sensitivity, specificity, and mean intersection-over-union were analysed using validation dataset. The accuracy was reported according to STARD-2015 statement.

RESULTS: A total of 567 intraoral photographs were collected and labelled, of which 80% were used for training and 20% for validation. Regarding training datasets, there were total 113,745,208 pixels with 9,270,413; 5,711,027; and 4,596,612 pixels were labelled as healthy, diseased, and questionable respectively. Regarding validation datasets, there were 28,319,607 pixels with 1,732,031; 1,866,104; and 1,116,493 pixels were labelled as healthy, diseased, and questionable, respectively. AI correctly predicted 1,114,623 healthy and 1,183,718 diseased pixels with sensitivity of 0.92 and specificity of 0.94. The mean intersection-over-union of the system was 0.60 and above the commonly accepted threshold of 0.50.