A novel semiautomatic segmentation protocol to evaluate guided bone regeneration outcomes: A pilot randomized, controlled clinical trial

Abstract Objectives The aims of this study were to (a) present a novel morphological contour interpolation (MCI) algorithm based method to evaluate grafted bone alterations following guided bone regeneration (GBR), (b) compare clinical and radiological outcomes of GBR with two different collagen membranes. Materials and methods The data were retrieved from an ongoing randomized controlled trial. Patients were randomly allocated into two groups: (a) control group (CG): Bio‐Gide (b) test group (TG): bovine dermis‐derived collagen membrane. Cone beam computed tomography examinations were performed 1 week (T0) and 6 months after surgery (T1). PES/WES at T1, grafted bone volume and density changes from T0 to T1 were recorded. Results Thirty‐six patients (16/20 in test/control group, respectively) were enrolled in the present study. Excellent inter‐observer reliability (ICC ≥ 0.97) was revealed for repeated measurements using this method. Significant volumetric reduction of grafted bone were found in both groups (test group: from 0.60 to 0.39 cm 3 , p < 0.01; control group: from 0.54 to 0.31 cm 3 , p < 0.01). Mean bone density (gray‐scale values) significantly increased from 305.12 to 456.69 in CG ( p < 0.01). In TG, it slightly increased from 304.75 to 393.27 ( p = 0.25). The mean PES/WES values were 13.84 (6.62/7.22) and 13.90 (6.70/7.20) for TG and CG, respectively. As for inter‐group comparison, no significant differences of grafted bone volume change, density change and PES/WES were found between two groups. Conclusion Within the limitations of this study, the novel MCI‐based method is a reproducible tool to segment and visualize changes of grafted bone in 3D. Furthermore, both collagen membranes could be used as a barrier membrane for GBR in humans.