BACKGROUND: The aim of this study was to explore possible causal relationships among several variables in the coronally advanced flap for root coverage procedure using structural learning of Bayesian networks. METHODS: Sixty consecutive patients with maxillary buccal recessions (>or=2 mm) were enrolled. All defects were treated with the coronally advanced flap procedure. Age, gender, smoking habits, recession depth, width of keratinized tissue, probing depth, distance between the incisal margin and the cemento-enamel junction, root sensitivity, and distance between the gingival margin and the cemento-enamel junction were recorded and calculated for all patients at baseline, immediately after surgery, and at 6 months after surgery. A structural learning algorithm of Bayesian networks was used. RESULTS: The distance between the gingival margin and the cemento-enamel junction immediately after surgery was affected by the baseline recession depth; deeper recessions were associated with a more apical location of the gingival margin after surgery. Moreover, complete root coverage also seemed to be affected by the location of the gingival margin after surgery; a more coronal location of the gingival margin after surgery was associated with a greater probability of complete root coverage. CONCLUSIONS: The use of structural learning of Bayesian networks seemed to facilitate the understanding of the possible relationships among the variables considered. The main result revealed that complete root coverage seemed to be influenced by the post-surgical position of the gingival margin and indirectly by the baseline recession depth.
BACKGROUND: The aim of this study was to explore possible causal relationships among several variables in the coronally advanced flap for root coverage procedure using structural learning of Bayesian networks. METHODS: Sixty consecutive patients with maxillary buccal recessions (>or=2 mm) were enrolled. All defects were treated with the coronally advanced flap procedure. Age, gender, smoking habits, recession depth, width of keratinized tissue, probing depth, distance between the incisal margin and the cemento-enamel junction, root sensitivity, and distance between the gingival margin and the cemento-enamel junction were recorded and calculated for all patients at baseline, immediately after surgery, and at 6 months after surgery. A structural learning algorithm of Bayesian networks was used. RESULTS: The distance between the gingival margin and the cemento-enamel junction immediately after surgery was affected by the baseline recession depth; deeper recessions were associated with a more apical location of the gingival margin after surgery. Moreover, complete root coverage also seemed to be affected by the location of the gingival margin after surgery; a more coronal location of the gingival margin after surgery was associated with a greater probability of complete root coverage. CONCLUSIONS: The use of structural learning of Bayesian networks seemed to facilitate the understanding of the possible relationships among the variables considered. The main result revealed that complete root coverage seemed to be influenced by the post-surgical position of the gingival margin and indirectly by the baseline recession depth.
Authors: Subhagata Chattopadhyay; Rima M Davis; Daphne D Menezes; Gautam Singh; Rajendra U Acharya; Toshio Tamura Journal: J Med Syst Date: 2010-10-13 Impact factor: 4.460