Regenerative treatment of periodontal defects was first published in the early 1980’s (Nyman et al. 1982), but the concept of a new periodontal attachment formation in intrabony periodontal defects was emphasized from the 1970’s (Melcher 1976). Since the first concepts on regenerative periodontal therapy, many different approaches have been introduced to achieve periodontal regeneration. The introduction of various biomaterials has made regenerative surgery more predictable and more straightforward. Applied surgical modalities and regenerative strategies are determined by the morphology and the extent of the intrabony defect. Decision-trees (Cortellini 2012), described in the literature provide treatment options for different clinical scenarios.
Defect morphology is determined by (i) direct clinical measurements (probing pocket depth: PPD, gingival recession: GR, clinical attachment loss: CAL) and (ii) two-dimensional (2D) radiographic images (intraoral radiographs: IR, and panoramic x-rays: PX). These tools are used for diagnostics and treatment planning of periodontally involved patients. The aforementioned methods are considered the gold standard in periodontal diagnostics, however there are a few drawbacks and in certain cases they don’t provide sufficient amount of information. Clinical studies have demonstrated, that clinicians constantly underestimated the extent of intrabony defects during direct clinical measurements (Eickholz et al. 1998, Vrotsos 1999, Christiaens 2018). IRs provide two-dimensional (2D) image, where overlapping anatomical structures make it difficult to accurately determine the true three-dimensional (3D) defect morphology (Eickholz et al. 1998, Christiaens 2018). Since the primary determining factors when selecting the regenerative treatment modality for intrabony periodontal defects are the morphology and the extent of the defect, without an accurate knowledge of these features, an exact treatment plan cannot be determined.
The application of cone-beam computed tomography (CBCT) in periodontal diagnostics diagnosis has been proposed by many authors (Mish et al. 2006; Kasaj et al. 2007; Walter et al. 2009). Series of in vitro and in vivo studies have demonstrated, that in certain cases CBCT is superior in the detection of periodontal defects (i.e. furcation defects, three wall intrabony defects, midbuccal intrabony defects and dehiscence type defects) then IRs (Vandenberghe et al. 2007; Vandenberghe 2008; Grimard et al. 2009, de Faria Vasconcelos 2012; Bagis et al. 2015; Cetmili et al. 2019) (Figure 1.), however it is difficult to justify the cost-benefit ratio of higher irradiation dose (Woebler et al. 2018; Walter et al. 2016). Therefore, CBCT should only be used for periodontal diagnosis, if conventional radiographic methods do not provide sufficient amount of information, which is in line with the 2017 recommendation of the American Academy of Periodontology (AAP) (Mandelaris et al. 2017).
CBCT images provide images in multiple orientations (sagittal, axial, coronal), however only one slice can be viewed at a time. DICOM (Digital Imaging and Communications in Medicine) imaging software automatically generated three-dimensional volume renders of datasets, but the reconstruction is done with basic threshold algorithms based on gray values of each voxel, and not based on anatomical structures. Due to the nature of cone beam tomography, artefacts and scattering, caused by teeth and metal restorations, compromise image quality (Queiroz et al. 2018), consequently visualization of small and delicate areas, such as the periodontium could be difficult.
Various surgical fields in general medicine such as: cardiac surgery, orthopedic surgery and cranio-maxillofacial surgery have utilized different radiographic image segmentation techniques to create patient specific digital three-dimensional anatomical renders and 3D printed models for diagnostics and treatment planning.
Aim of this study is to present a method for 3D visualization of intrabony periodontal defects with the help of virtual, patients specific models reconstructed from CBCT datasets and to evaluate the accuracy of the models, by comparing the results with direct intrasurgical measurements.