Machine Learning Based Software to Predict Type of Gingival Recession Surgery

Abstract

ABSTRACT Purpose: The goal of this study is to identify the most important variables affecting gingival recession and to develop a machine learning-based software using these variables. Materials & Methods: 132 mandibular/maxillary right/left teeth #1, #2, #3, #4, and #5 were included in the study. Recession depth, recession width, width of keratinized gingiva, plaque index, buccogingival tissue thickness, frenulum position and mobility were recorded before and 6-month after surgery. One of the following methods has been selected: Gingival unit graft, coronally advanced flap, coronally advanced flap+connective tissue graft, and coronally advanced flap+platelet-rich fibrin. Software was developed to predict the type of gingival recession surgery. Results: While the gingival unit graft group had the highest pre- and post-recession depth values, the coronally advanced flap group had the highest pre-recession width, pre-width of the keratinized gingiva and post-width of the keratinized gingiva values. A significant difference was found between surgical type categories for all variables except gender, post-buccogingival tissue thickness and post-frenulum position (p<0.05). Random Forest was found to be the best performing method both for surgery categories and overall based on accuracy and F-measure. Accuracy value was found 90.0% for gingival unit graft, 62.5% for coronally advanced flap, 71.4% for coronally advanced flap+connective tissue graft, and 97.8% for coronally advanced flap+platelet-rich fibrin. Conclusions: The machine learning software could evaluate the data accumulated in the database using the decision trees method and predict the prognosis of surgical techniques to treat gingival recession. The software developed will help physicians determine the optimal treatment approach. Key Words: Gingival recession, Mucogingival surgery, Decision trees, Machine learning, Predictive decision model.

Keywords

• Gingival recession
• Mucogingival surgery
• Decision trees
• Machine learning
• Predictive decision model

References

1. Cortellini P, Bissada NF. Mucogingival conditions in the natu- ral dentition: Narrative review, case definitions, and diagnos- tic considerations. J Periodontol. 2018;89 Suppl 1:S204–s213. doi:10.1002/jper.16-0671.
2. 2. Guttiganur N, Aspalli S, Sanikop MV, Desai A, Gaddale R, De- vanoorkar A. Classification systems for gingival recession and suggestion of a new classification system. Indian J Dent Res. 2018;29(2):233–237. doi:10.4103/ijdr.IJDR_207_17.
3. 3. Kim DM, Neiva R. Periodontal soft tissue non-root cover- age procedures: a systematic review from the AAP Regen- eration Workshop. J Periodontol. 2015;86(2 Suppl):S56–572. doi:10.1902/jop.2015.130684.
4. 4. Cairo F, Pagliaro U, Nieri M. Treatment of gingival recession with coronally advanced flap procedures: a systematic review. J Clin Periodontol. 2008;35(8 Suppl):136–162. doi:10.1111/j.1600- 051X.2008.01267.x.
5. 5. Chambrone L, Tatakis DN. Periodontal soft tissue root cov- erage procedures: a systematic review from the AAP Regen- eration Workshop. J Periodontol. 2015;86(2 Suppl):S8–51. doi:10.1902/jop.2015.130674.
6. 6. Kassab MM, Cohen RE. The etiology and prevalence of gingival recession. J Am Dent Assoc. 2003;134(2):220–225. doi:10.14219/jada.archive.2003.0137.
7. 7. Cairo F, Nieri M, Cincinelli S, Mervelt J, Pagliaro U. The in- terproximal clinical attachment level to classify gingival re- cessions and predict root coverage outcomes: an explorative and reliability study. J Clin Periodontol. 2011;38(7):661–666. doi:10.1111/j.1600-051X.2011.01732.x.
8. 8. Holbrook T, Ochsenbein C. Complete coverage of the denuded root surface with a one-stage gingival graft. Int J Periodontics Restorative Dent. 1983;3(3):8–27.

9. 9. Miller J P D. Root coverage grafting for regeneration and aes- thetics. Periodontol 2000. 1993;1:118–127. doi:10.1111/j.1600- 0757.1993.tb00213.x.
10. 10. Agudio G, Nieri M, Rotundo R, Cortellini P, Pini Prato G. Free gingival grafts to increase keratinized tissue: a retrospective long-term evaluation (10 to 25 years) of outcomes. J Periodontol. 2008;79(4):587–594. doi:10.1902/jop.2008.070414.
11. 11. Albandar JM, Kingman A. Gingival recession, gingival bleed- ing, and dental calculus in adults 30 years of age and older in the United States, 1988-1994. J Periodontol. 1999;70(1):30–43. doi:10.1902/jop.1999.70.1.30.
12. 12. Kan JY, Morimoto T, Rungcharassaeng K, Roe P, Smith DH. Gingival biotype assessment in the esthetic zone: visual ver- sus direct measurement. Int J Periodontics Restorative Dent. 2010;30(3):237–243.
13. 13. Mlinek A, Smukler H, Buchner A. The use of free gingi- val grafts for the coverage of denuded roots. J Periodontol. 1973;44(4):248–254. doi:10.1902/jop.1973.44.4.248.
14. 14. Wennström JL, Zucchelli G. Increased gingival dimensions. A significant factor for successful outcome of root coverage pro- cedures? A 2-year prospective clinical study. J Clin Periodontol. 1996;23(8):770–777. doi:10.1111/j.1600-051x.1996.tb00608.x.
15. 15. Zucchelli G, Testori T, De Sanctis M. Clinical and anatomical factors limiting treatment outcomes of gingival recession: a new method to predetermine the line of root coverage. J Peri- odontol. 2006;77(4):714–721. doi:10.1902/jop.2006.050038.
16. 16. Han J, Kamber M, Pei J. Concepts and Techniques. Morgan Kaufmann; 2011.
17. 17. Francis NK, Luther A, Salib E, Allanby L, Messenger D, Alli- son AS, et al. The use of artificial neural networks to predict delayed discharge and readmission in enhanced recovery fol- lowing laparoscopic colorectal cancer surgery. Tech Coloproctol. 2015;19(7):419–428. doi:10.1007/s10151-015-1319-0.
18. 18. Tanaka T, Voigt MD. Decision tree analysis to stratify risk of de novo non-melanoma skin cancer following liver trans- plantation. J Cancer Res Clin Oncol. 2018;144(3):607–615. doi:10.1007/s00432-018-2589-5.
19. 19. Kantardzic M. Data mining: concepts, models, methods, and algorithms. John Wiley & Sons; 2011.
20. 20. Silness J, Loe H. Periodontal Disease In Pregnancy. II. Correlation Between Oral Hygiene And Periodon- tal Condtion. Acta Odontol Scand. 1964;22:121–135. doi:10.3109/00016356408993968.
21. 21. Malpartida-Carrillo V, Tinedo-Lopez PL, Guerrero ME, Amaya- Pajares SP, Özcan M, Rösing CK. Periodontal phenotype: A review of historical and current classifications evaluating dif- ferent methods and characteristics. J Esthet Restor Dent. 2021;33(3):432–445. doi:10.1111/jerd.12661.
22. 22. Miller SC. Successful Treatment of Periodontal Disease. The Military Surgeon (United States). 1951;108(3):191–193.
23. 23. Aroca S, Keglevich T, Barbieri B, Gera I, Etienne D. Clinical eval- uation of a modified coronally advanced flap alone or in combi- nation with a platelet-rich fibrin membrane for the treatment of adjacent multiple gingival recessions: a 6-month study. J Pe- riodontol. 2009;80(2):244–252. doi:10.1902/jop.2009.080253.
24. 24. Bittencourt S, Del Peloso Ribeiro E, Sallum EA, Sallum AW, Nociti J F H, Casati MZ. Comparative 6-month clinical study of a semilunar coronally positioned flap and subepithelial con- nective tissue graft for the treatment of gingival recession. J Periodontol. 2006;77(2):174–181. doi:10.1902/jop.2006.050114.
25. 25. Kuru B, Yıldırım S. Treatment of localized gingival recessions using gingival unit grafts: a randomized controlled clinical trial. J Periodontol. 2013;84(1):41–50. doi:10.1902/jop.2012.110685.
26. 26. Paolantonio M, di Murro C, Cattabriga A, Cattabriga M. Sub- pedicle connective tissue graft versus free gingival graft in the coverage of exposed root surfaces. A 5-year clinical study. J Clin Periodontol. 1997;24(1):51–56. doi:10.1111/j.1600- 051x.1997.tb01184.x.
27. 27. Paolantonio M, Femminella B, Coppolino E, Sammartino G, D’Arcangelo C, Perfetti G, et al. Autogenous periosteal bar- rier membranes and bone grafts in the treatment of peri- odontal intrabony defects of single-rooted teeth: a 12-month reentry randomized controlled clinical trial. J Periodontol. 2010;81(11):1587–1595. doi:10.1902/jop.2010.100094.
28. 28. Pini-Prato GP, Cairo F, Nieri M, Franceschi D, Rotundo R, Cortellini P. Coronally advanced flap versus connective tis- sue graft in the treatment of multiple gingival recessions: a split-mouth study with a 5-year follow-up. J Clin Periodontol. 2010;37(7):644–650. doi:10.1111/j.1600-051X.2010.01559.x.
29. 29. Pini Prato GP, Baldi C, Nieri M, Franseschi D, Cortellini P, Clauser C, et al. Coronally advanced flap: the post-surgical position of the gingival margin is an important factor for achiev- ing complete root coverage. J Periodontol. 2005;76(5):713–722. doi:10.1902/jop.2005.76.5.713.
30. 30. Pini Prato G, Pagliaro U, Baldi C, Nieri M, Saletta D, Cairo F, et al. Coronally advanced flap procedure for root coverage. Flap with tension versus flap without tension: a randomized controlled clinical study. J Periodontol. 2000;71(2):188–201. doi:10.1902/jop.2000.71.2.188. Machine Learning Based Software for Gingival Recession Surgery | 9
31. 31. Gregori D, Petrinco M, Bo S, Rosato R, Pagano E, Berchialla P, et al. Using data mining techniques in monitoring diabetes care. The simpler the better? J Med Syst. 2011;35(2):277–281. doi:10.1007/s10916-009-9363-9.
32. 32. Karaolis MA, Moutiris JA, Hadjipanayi D, Pattichis CS. Assess- ment of the risk factors of coronary heart events based on data mining with decision trees. IEEE Trans Inf Technol Biomed. 2010;14(3):559–566. doi:10.1109/titb.2009.2038906.
33. 33. Liu H, Xie G, Mei J, Shen W, Sun W, Li X. An efficacy driven approach for medication recommendation in type 2 diabetes treatment using data mining techniques. Stud Health Technol Inform. 2013;192:1071. doi:10.3233/978-1-61499-289-9-1071.
34. 34. Page RC, Krall EA, Martin J, Mancl L, Garcia RI. Va- lidity and accuracy of a risk calculator in predicting peri- odontal disease. J Am Dent Assoc. 2002;133(5):569–576. doi:10.14219/jada.archive.2002.0232.
35. 35. Page RC, Martin J, Krall EA, Mancl L, Garcia R. Longitu- dinal validation of a risk calculator for periodontal disease. J Clin Periodontol. 2003;30(9):819–827. doi:10.1034/j.1600- 051x.2003.00370.x.
36. 36. Persson GR, Mancl LA, Martin J, Page RC. Assessing periodontal disease risk: a comparison of clinicians’ assessment versus a computerized tool. J Am Dent Assoc. 2003;134(5):575–582. doi:10.14219/jada.archive.2003.0224.
37. 37. Covani U, Marconcini S, Derchi G, Barone A, Giacomelli L. Relationship between human periodontitis and type 2 dia- betes at a genomic level: a data-mining study. J Periodontol. 2009;80(8):1265–1273. doi:10.1902/jop.2009.080671.
38. 38. Fan J, Nunn ME, Su X. Multivariate Exponential Survival Trees And Their Application to Tooth Prognosis. Comput Stat Data Anal. 2009;53(4):1110–1121. doi:10.1016/j.csda.2008.10.019.
39. 39. Podgorelec V, Kokol P, Stiglic B, Rozman I. Decision trees: an overview and their use in medicine. J Med Syst. 2002;26(5):445– 463. doi:10.1023/a:1016409317640.
40. 40. Wolfenden L, Williams CM, Kingsland M, Yoong SL, Nathan N, Sutherland R, et al. Improving the impact of public health service delivery and research: a decision tree to aid evidence- based public health practice and research. Aust N Z J Public Health. 2020;44(5):331–332. doi:10.1111/1753-6405.13023.
41. 41. Aroca S, Barbieri A, Clementini M, Renouard F, de Sanctis M. Treatment of class III multiple gingival recessions: Prognostic factors for achieving a complete root coverage. J Clin Periodon- tol. 2018;45(7):861–868. doi:10.1111/jcpe.12923.
42. 42. Kuis D, Sciran I, Lajnert V, Snjaric D, Prpic J, Pezelj-Ribaric S, et al. Coronally advanced flap alone or with connective tissue graft in the treatment of single gingival recession de- fects: a long-term randomized clinical trial. J Periodontol. 2013;84(11):1576–1585. doi:10.1902/jop.2013.120451.
43. 43. Borghetti A, Gardella JP. Thick gingival autograft for the cover- age of gingival recession: a clinical evaluation. Int J Periodontics Restorative Dent. 1990;10(3):216–229.
44. 44. Remya V, Kishore Kumar K, Sudharsan S, Arun KV. Free gingival graft in the treatment of class III gingival recession. Indian J Dent Res. 2008;19(3):247–252. doi:10.4103/0970-9290.42959.
45. 45. Souza SL, Macedo GO, Tunes RS, Silveira e Souza AM, Novaes J A B, Grisi MF, et al. Subepithelial connective tissue graft for root coverage in smokers and non-smokers: a clinical and histologic controlled study in humans. J Periodontol. 2008;79(6):1014– 1021. doi:10.1902/jop.2008.070479.
46. 46. Huang LH, Neiva RE, Wang HL. Factors affecting the outcomes of coronally advanced flap root coverage procedure. J Periodontol. 2005;76(10):1729–1734. doi:10.1902/jop.2005.76.10.1729.
47. 47. Müller HP, Eger T, Schorb A. Gingival dimensions after root coverage with free connective tissue grafts. J Clin Periodontol. 1998;25(5):424–430. doi:10.1111/j.1600-051x.1998.tb02466.x.
48. 48. Ercan E, Bilgin E, Koprucu S, Kayipmaz S. Evaluation of the relationship between gingival phenotype and alveolar bone morphology. Mucosa. 2019;2(1):6–13.
49. 49. Mirko P, Miroslav S, Lubor M. Significance of the labial frenum attachment in periodontal disease in man. Part I. Classification and epidemiology of the labial frenum attachment. J Periodontol. 1974;45(12):891–894. doi:10.1902/jop.1974.45.12.891.
50. 50. Bernimoulin J, Curilovié Z. Gingival recession and tooth mo- bility. J Clin Periodontol. 1977;4(2):107–114. doi:10.1111/j.1600- 051x.1977.tb01890.x.
51. 51. Kerry GJ, Morrison EC, Ramfjord SP, Hill RW, Caffesse RG, Nissle RR, et al. Effect of periodontal treatment on tooth mobility. J Periodontol. 1982;53(10):635–638. doi:10.1902/jop.1982.53.10.635.