Digital Assessment of Gingival Dimensions of Healthy Periodontium

The aim of the present study was to re-visit the gingival dimension using digital scanning in a healthy Korean population. Forty-eight periodontally healthy volunteers (38 males and 10 females, mean age: 24.3 ± 2.2 years) were included. The mucogingival junction was highlighted using 2.5% diluted iodine solution. Then, the facial gingiva and mucosa of both jaws were digitally scanned using an intraoral digital scanner. Using computer software and periodontal probing, the heights and areas of keratinized gingiva (KG) and attached gingiva (AG) were measured. Similar distribution patterns in the gingival heights were noted in the maxilla and mandible. The maxilla showed substantially greater gingival values than the mandible. The heights of the KG and AG were notably smaller on the mandibular first premolar (2.37 mm and 1.07 mm, median value) and second molar (3.28 mm and 1.78 mm) than on the other teeth. The area of the KG was the largest in the canine (63.74 mm2 and 46.85 mm2) and first molar (64.14 mm2 and 58.82 mm2) in each jaw. Mandibular first and second molars, mandibular canine, and maxillary canine showed the highest value of the area under the receiver operation characteristics curve (>0.7) for differentiating between males and females. The gingival dimensions recorded using intraoral scanner demonstrated similar distribution patterns as in previous studies.


Introduction
Several studies have investigated the dimensions of the keratinized gingiva (KG) and attached gingiva (AG) around natural dentition [1][2][3][4], in pursuit of obtaining anatomic information and finding adequate gingival dimension for periodontal and mechanical stability. Especially, some surgical procedures, such as bone denudation procedure and apically positioned flap surgery, were once proposed for the latter purpose [5,6].
Literature demonstrated heterogeneity regarding the role of KG/AG. A representative study by Lang and Löe demonstrated that 2 mm of KG (1 mm of AG) was sufficient for periodontal health in the presence of professional maintenance [2]. On the contrary, others published later failed to show no specific relationship between KG/AG and periodontal health [3,7], leading to a consensus that there is no adequate gingival dimension for periodontal health. However, recent evidence suggests a substantial role of the gingival dimension as a part of periodontal phenotype to prevent gingival recession [8,9].
In terms of understanding the esthetics and pathophysiology of the periodontium and adjacent structures, dental clinicians must have an anatomical overview of the periodontium, such as dimension, tendency, and composition. From histological and morphological leading to a clear demarcation between the KG and alveolar mucosa (mucogingival junction [MGJ]). Then, an intraoral digital scanner (i500, iScan version 1.2.0.1; Medit, Seoul, Korea) was used for whole arch scanning. The extent of the scanning was up to the deepest portion of the vestibule. The same scanning procedure was used for all patients: The KG around the anterior region was first scanned. Then, the upper and the lower-left posterior regions were scanned with the examiner standing in the 12 o'clock position to the patient. Then, in the 7 o'clock position, the examiner scanned the upper and the lower-right posterior regions. The scanning procedure was conducted under the recommended conditions of level-1 filtering and a 21.0-mm scan depth ( Figure 1A).

Clinical Measurements
With a UNC-15 probe, the probing depth, plaque index [17], and gingival index [18] were measured. Plaque and gingival indices were recorded prior to intraoral scanning. After scanning, the probing depth was measured.

Analysis
After the digital scans were acquired, the scanned data in the polygon file format (PLY) were exported from the scanning program (Medit link, Medit). The dataset imported into the image analysis software (Geomagic Design X™, 3D Systems, Inc., Rock Hill, CA, USA) were visualized in 3D reconstructed images ( Figure 1B). A single investigator (J.L.) who was trained in using the software calibrated it by measuring 10 random

Clinical Measurements
With a UNC-15 probe, the probing depth, plaque index [17], and gingival index [18] were measured. Plaque and gingival indices were recorded prior to intraoral scanning. After scanning, the probing depth was measured.

Analysis
After the digital scans were acquired, the scanned data in the polygon file format (PLY) were exported from the scanning program (Medit link, Medit). The dataset imported into the image analysis software (Geomagic Design X™, 3D Systems, Inc., Rock Hill, CA, USA) were visualized in 3D reconstructed images ( Figure 1B). A single investigator (J.L.) who was trained in using the software calibrated it by measuring 10 random samples two times each in a span of one week under the supervision of a senior investigator (J.C.P.). All measurements were rounded to the nearest 0.01 mm. The calibration session yielded an intraclass correlation coefficient (ICC) of 0.90.

Height of Keratinized Gingiva
For each tooth position, the height of the KG was greater in the maxilla than in the mandible. The greatest and the least differences were observed in the second molar (4.56 mm (3.98, 5.23) in the maxilla vs. 2.90 mm (2.51, 3.10) in the mandible; data are presented as median value with 95% confidence interval) and in the second premolar (3.95 mm (3.62, 4.18) vs. 3.28 mm (3.15, 3.52)), respectively (Table 1, Figure 2).

Results
Of the 51 volunteers (38 males and 13 females, mean age: 24.2 ± 2.1), three had missing teeth, so their data were excluded. Finally, 48 volunteers (38 males and 10 females, mean age: 24.3 ± 2.2) were included in the analysis. The data on tooth type and tooth number in both jaws are presented in Tables 1-3 and Tables A1-A3, respectively.

Height of Keratinized Gingiva
For each tooth position, the height of the KG was greater in the maxilla than in the mandible. The greatest and the least differences were observed in the second molar (4.56 mm (3.98, 5.23) in the maxilla vs. 2.90 mm (2.51, 3.10) in the mandible; data are presented as median value with 95% confidence interval) and in the second premolar (3.95 mm (3.62, 4.18) vs. 3.28 mm (3.15, 3.52)), respectively (Table 1, Figure 2).  In both jaws, similar trends of KG distribution were observed. The least value was noted in the first premolar. The values in this area were 3.39 mm (2.95, 3.75) in the maxilla and 2.37 mm (2.18, 2.55) in the mandible. The height of the KG in the central and lateral incisors was comparable in each jaw (5.17 mm (4.64, 5.51) and 5.33 mm (4.80, 5.87) in the maxilla, 4.18 mm (4.01, 4.37) and 4.52 mm (4.25, 4.80) in the mandible) and was higher than that in other areas of each jaw. In the maxilla, the height of the KG in the first (4.63 mm (4.04, 4.99)) and the second molars was similar (4.56 mm (3.98, 5.23)), but in the mandible, the value was greater in the first molar (3.78 mm (3.60, 3.86)) than in the second molar (2.90 mm (2.51, 3.10)). The height of the KG was not statistically significantly different between male and female volunteers (p > 0.05).

Height of Attached Gingiva
Same as the height of the KG, for each tooth position, the height of the AG was greater in the maxilla than in the mandible. In both jaws, the first premolar showed the lowest height (1.85 mm (1.38, 2.12) vs. 1.07 mm (0.79, 1.38)). In the maxilla, the median value of the AG height for all the teeth was over 2 mm except for the first premolar. However, in the mandible, the height in the first and second premolars (1.78 mm (1.45, 1.97)) and second molars (1.42 mm (1.22, 1.71)) was less than 2 mm (Table 2, Figure 2). Height of the AG was not statistically significantly different between male and female volunteers (p > 0.05).

Area of Keratinized Gingiva
In both jaws, the area of the KG was the largest in the first molar (64. 14 Figure 2). The area of the KG in the maxillary and mandibular canines and molars was statistically significantly different between male and female volunteers (p < 0.05).

Determination of Sex on a Basis of Gingival Dimensions
AUCs were calculated for the parameters of all teeth. Then, the parameters with AUC over 0.7 were selected as per the recommendations of Muller et al. (2005) [19]. In their study, 1.0-0.9 AUC was considered excellent, 0.9-0.8, good, 0.8-0.7, fair, 0.7-0.6, poor, and 0.6-0.5, fail. The selected parameters were area of the KG in the mandibular second molar (AUC = 0.74, cut-off value = 31.56 mm 2 ), mandibular canine (AUC = 0.73, cut-off value = 40.47 mm 2 ), maxillary canine (AUC = 0.72, cut-off value = 64.44 mm 2 ), and mandibular first molar (AUC = 0.71, cut-off value = 53.1 mm 2 ) ( Table 4). For example, when the area of the KG on the mandibular second molar was over 31.56 mm 2 , the subject was likely to be male. The sensitivity, specificity, and accuracy are presented in Table 4.

Discussion
The gingiva has a convex curvilinear shape. In the past, this curvilinear outline could not be recorded the way it is; the measurements from a straight line connecting two landmarks were used to determine a representative value. Until now, no quantitative information regarding curvilinear gingival line was reported. Considering this, the present study investigated the gingival dimensions of periodontally healthy young adults with the use of an intraoral scanner. The current investigation can help to enhance anatomic information with high accuracy.
In general, the distribution pattern of the gingival height in the present study was in conformity to previous studies. A comparison of the maxilla and mandible revealed that the height of the KG and AG were higher in maxillary teeth than in the mandibular teeth in the same position. The height of the KG and AG in both jaws showed similar fluctuation trends by tooth type (See Figure 2). Regarding the maxillary central and maxillary lateral incisors, little heterogeneity was found in the gingival distribution pattern between different studies (including the present study). In the present study, the gingival heights in the maxillary central incisor (5.17 mm for KG, 4.35 mm for AG; median value) showed different tendency from those in the maxillary lateral incisor (5.33 mm for KG, 3.98 mm for AG). This tendency (depending on KG or AG or both) is in line with the findings of the study by Adesola et al. [20] but not of Bower [4], Ainamo and Loë [1] and Shirmohammadi, Faramarzi, Lafzi [21] (see the values of these studies in Table A4).
In both jaws, the lowest KG height was reported in the first premolar (3.39 mm in the maxilla, 2.37 mm in the mandible), which resulted in the low height of the AG in this position. Especially, the height of AG in the mandibular second premolar was approximately 1 mm, which is a striking finding compared to that of other studies performed in different countries (1.8-2.42 mm) (See Table A4) [1,2,4,[20][21][22]. Frenular attachment adjacent to the premolar appears to affect the height of the AG.
Interestingly, the height of AG in the mandibular second molar was also small (approximately 1.0 mm) in the present study. Anatomically, the external oblique ridge starts from third molar area, leading to the extension of the buccal mucosa to the mandibular second molar. Owing to this, the height of the KG in the mandibular second molar tends to be smaller than that in the mandibular first molar and maxillary second molar, and the height of the AG changes concomitantly. It is also noteworthy that the height of the AG in the mandibular second molar of this study was lower compared to other studies [1,2,4,[20][21][22]. This difference might be attributable to the difference in jaw growth and mucosal attachment.
The area of the KG in each tooth position was firstly reported in the present study. The different mesio-distal widths of each tooth considerably affected the area of the KG. For example, the canine and the first molar had larger KG area than the other teeth. The area of the KG may influence the mechanical stability of the soft tissue, such as susceptibility to the recession. However, epidemiologic studies demonstrated heterogeneous findings [23][24][25].
The heights of the KG and AG in the Korean population are smaller than the values reported in other studies [1,2,4,[20][21][22]. Previously, the KG and/or AG were measured in the US [4], Denmark [1], Iran [21], India [22], Nigeria [20]. Especially, the height of the AG in the Iranian population was notably greater than that in other populations [21]. This indicates ethnic and racial differences in the dimensions of the KG and AG, which was also discussed in other studies [20,21]. Other reasons behind this difference may be as follows: (1) The age groups targeted were different across the studies. Previously, an increase in the AG was found with age [26], owing to the constant position of the MGJ and attrition-related tooth eruption. (2) Digital measurements provide different levels of accuracy as opposed to conventional measurements. A recent study comparing histological and digital/clinical measurements in a pig jaw model showed that the digital technique had superior accuracy and reliability in measuring the KG than the periodontal probe [15]. (3) Periodontal situation, a history of orthodontic treatment, tooth malposition was inconsistent or not reported.
A systematic review investigating the correlation between periodontal phenotype and gingival dimensions showed that the mean KG height ranged between 2.75 and 5.44 mm in the thin phenotype and between 5.09 mm and 6.65 mm in the thick phenotype [27]. According to this, most of the present study subjects might fall under the thin phenotype, which may be in line with recent evidence that Asian population has a thin gingival phenotype compared to white population [9]. However, this speculation should be back up by further data from other measurements, such as cone-beam computed tomography, ultrasonography, or endodontic instruments [28].
In the present study, we tried to differentiate between male and female volunteers using the dimensional data. AUC > 0.7 was reported for the area of the KG in the mandibular second molar, mandibular canine, maxillary canine, and mandibular first molar. Among those, the mandibular second molar showed the highest sensitivity, and the mandibular canine and first molar showed the highest specificity. The dataset was further used for machine learning tools, such as random forest and Xgboost, despite the small sample size. However, some of the data from female volunteers were spent in the training set, which led to less data in the test set. To draw more reliable results from machine learning, a larger dataset should be collected from both male and female.
There are some limitations and considerations in the present study. First, gingival phenotype was not recorded, as mentioned above. The addition of such data would provide more comprehensive information about the gingival tissues. Second, there was a discrepancy between the numbers of male and female subjects, which might affect statistical results. Third, the pattern of gingival dimension in the present study was not different from that in the previous studies, indicating traditional measurement is still valid when high inter/ intra-examiner reproducibility and reliability are obtained.
Several things should be covered in future studies, such as comparing between measurements using a periodontal probe and using digital scanning, gingival dimensions with/without periodontal diseases, and gingival dimension change over time using superimposition. Especially, the data from digital scanning technology can be utilized to visualize the change for the area of the investigators' interest and measure profilometric change in great detail, which was not feasible using traditional methods.

Conclusions
In conclusion, the gingival dimensions recorded using an intraoral scanner showed a similar distribution pattern as that in previous studies. However, the dimensions appear to be influenced by race and/or ethnicity, especially in the mandibular canine and second molar. Some of the gingival dimensions may be used for differentiating between male and female.  Informed Consent Statement: Written informed consent has been obtained from the patient(s) to publish this paper.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to patients' privacy.

Acknowledgments:
The authors appreciate the research team at the Department of Periodontology, College of Dentistry Dankook University.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.