spinner
Geistlich Blog header image
3D imaging

A more accurate measure of soft tissue changes

08.01.2019
share this article

Quantifying treatment outcomes is the first step towards continuous improvement. A new technique allows accurate and precise measurement of the increase in soft tissue thickness.

Adjunct Prof. Dr. Thiago Morelli | United States

Traditionally, visual inspection1 and periodontal probes2 have been used to assess tissue transparency and classify gingival biotype. However, it has been shown that visual assessment is associated with misclassification of approximately half of thin-scalloped cases.3 The use of CBCT has been tested and shown to provide more reliable measures.4,5 However, this method also has several drawbacks, including the limitation of linear measurements only, scattering effect that can affect analysis and additional patient exposure to radiation for comparative analysis.


3-D images from digital optical scanning

Digital optical scanning and assessment methods have been introduced to measure volume changes of oral tissues over time6,7 and provide a new perspective for measuring and quantifying soft tissue volume longitudinally after reconstruction or regenerative periodontal procedures.
Calibration preclinical and clinical studies demonstrate precision and reliability of this non-invasive method to assess soft tissue volume changes.8,9 Advances in digital optical scanning quality and precision and the use of non-contact, reverse engineering software have provided the potential to not only precisely measure soft tissue changes but also changes in 3-dimensional (3-D) images, providing an additional and visually more comprehensible perspective for clinical research analysis.


Three techniques for superimposition

Researchers in the field of regenerative medicine have always been interested in quantifying the effect of treatment on craniofacial morphology. In recent years, 3-D imaging techniques have been used widely in maxillofacial surgery, dental implantology and various other medical disciplines. The 3-D model superimposition can help identify treatment goals, identify ideal treatment modalities, predict treatment results and evaluate treatment outcomes. Various techniques have been reported for superimposition of 3-D datasets derived from conventional computed tomography, CBCT, or intra-oral optical scanners. They include:

  • Landmark-based superimposition
  • Surface-based superimposition
  • Voxel-based superimposition.

The validity of the first two superimposition techniques depends on the accuracy of landmark identification and the precision of the 3-D surface models.
In most software, the 3-D differences of the superimposed models are translated into 2D color codes that represent the distance between corresponding points. However, the quantification of the structural change is not always straightforward.


A new ­technique based on STL data

Recently our group developed a technique for superimposing 3-D models and accurately quantifying the volume changes after regenerative bone and soft tissue procedures. The technique involves the acquisition of 3-D surface models that can be obtained from CBCT or intra-oral optical scanners.
When utilizing a CBCT dataset, it is necessary to translate the DICOM file originated from the CBCT file to a Standard Tesselation Language (STL) file. For accurate 3-D superimposition, it is mandatory that the files used for superimposition be taken in the same CBCT unit with identical setup. Another important and critical component is the presence of metal in the area to be evaluated. The scatter effect can potentially modify the morphological structure of the 3-D reconstructed model, limiting the precision of the superimposition and ultimately the measurement and quantification of the volume change.
The intra-oral scanner we have used to obtain 3-D reconstructed images for soft tissue analysis is Trios 3® (3-Shape, Copenhagen, Denmark). Trios 3® is a powerful, extremely fast and light scanner. It works under the principle of confocal microscopy and ultrafast optical scanning. It is powder-free and produces high-quality, in-color (true color) images. In a validation study, Imburgia and collaborators reported that Trios 3® showed a trueness value ranging from 50.2 µm to 67.2 µm and a precision value ranging from 24.5 µm to 31.5 µm.10

A triangular representation of 3-D surface geometry

Surface-based registration could qualify as a valid alternative to the commonly used voxel-based approaches used in medical imaging. STL is an open-source, surface-based format, like DICOM for voxel data, and it is easily accessible through most commercial and freeware software applications. Such surface models have been widely used in industry, particularly in engineering and architecture, for rapid prototyping and computer-aided manufacturing. These 3-D datasets allow for easy information exchange and communication among scientists.
Surface models do not contain volume data, but instead use 3-D surface data that are different from the data obtained from a CBCT or intra-oral scanner – a triangular representation of 3-D surface geometry.
Also, in this case, the form of data, the surface preprocessing (e.g., smoothing, segmentation), the transformation model and the choice of reference structures should be considered as potential sources of error when superimposing surface models. In a recent study, Gkantidis et al. evaluated 3-D superimposition techniques on various skeletal structures using surface models and concluded that it can provide accurate, precise and reproducible results.11


Measuring volume changes

For surface and volume analysis, we used the Geomagic Qualify® software (Raindrop Geomagic, Research Triangle Park, NC, USA) modified to calculate total volume. A precision analysis study using CAD files determined that surface reconstruction using Geomagic Qualify® software provided a reliable analysis with a maximum deviation of 0.06 mm, standard deviation of 0.003 mm and an average error of 0.002 mm.12
Using 3-D models reconstructed from the intra-oral scanner, our average error was 0.04 mm and 0.07 mm, when using 3-D models reconstructed from the CBCT dataset. We used the three-point registration method as a reference point to superimpose the 3-D surface models. Once superimposed, 3-D analysis was performed, and the distances between corresponding areas were color-coded on the superimposed models for visualization.
The distance between two points can be calculated at any selected point, giving the linear measurement described in most studies using 3-D surface analysis. In order to calculate volume, it is necessary to pre-select the area in the reference model and replicate it into the tested model. Once 3-D analysis is complete, the software algorithm can generate the specific volume difference between the reference and test model.


Conclusion

Advances in digital tools used to obtain 3-D datasets can provide accurate, non-invasive and visually instructive methodologies for diagnosis, treatment planning and evaluation of the efficacy of treatment options, providing the clinician with valuable information for decision-making.
Most importantly, the use of digital technology for the evaluation of treatment outcomes can potentially demonstrate not only the statistical significance between different treatment options but also clinical significance, by visualizing the total volume gain or loss after surgical procedures.

 

 

Adjunct Prof. Dr. Thiago Morelli

Adjunct Prof. Dr. Thiago Morelli | United States

Private practice in Raleigh
University of North California
School of Dentistry, Chapell Hill

Your Comment

 Yes, I have read the privacy policy and agree to it.