gms | German Medical Science

In conventional operation planning for cranio-maxillofacial surgery, cephalometric measurements, i.e. measurements of the dimensions are taken by hand on lateral X-ray shots.

The goal of this work is to develop an interactive graphical software application for 3D cephalometry.

Main benefit of the computer approach is the ability to view the anatomy from different viewpoints and using multiple zoom factors.

In addition, working directly in 3D avoids planning inaccuracies resulting from imprecise correlation of two dimensional features observed on the X-ray projections to their anatomical counterpart.

The first step in conventional planning consists in taking a photograph of the patient and a lateral X-ray shot. Next, the surgeon delineates skin and skull profiles on a semi-transparent sheet. The proper cephalometric analysis is then performed by identifying a set of landmarks and measuring distances and angles between them.

This data serves as a base for the decision of the bone movement that will be executed in the operations room. Prediction of the outcome of the displacement is done by cutting the profile sketches and moving them by hand. Steps of the conventional planning procedure can be seen in Figure 1 [Fig. 1] Conventional Planning Method.

Main limitations of the conventional planning techniques are the following:

• The focus lies on a particular aspect of the procedure (e.g. dental occlusion, 2D profile) and a certain integration effort is required to obtain the global view of the planned intervention.
• This integration step generates additional imprecision in the plan (e.g. in the transfer of displacement values).
• Taking lateral asymmetries of the face into account when working on a 2D profile is difficult and inaccurate.
• Imprecise correlation of two-dimensional features observed on the X-ray projections to their three-dimensional counterpart generates inaccuracies in the plan.

The first step with the computer approach is to make an X-ray tomography (CT) scan of the patient’s head. From this volumetric image dataset, a 3D surface model of the patient’s skull is generated (Figure 2 [Fig. 2]).

This model is then loaded into the cephalometry analyzer module, which allows the surgeon to assess and quantify anatomical deformities and asymmetries, and to define landmarks, lines, distances, angles and reference planes in three dimensions.

During relocation of the bone segments, cephalometric data is adapted in real-time, thus immediately showing the effect of the movement onto the skulls profile.

Visualization is improved with the ability to view the anatomy from different viewpoints and zoom factors.

Inaccuracies resulting from imprecise correlation of two dimensional features observed on the X-ray projections to their anatomical counterpart can be avoided by working directly in 3D.

3D visualization is particularly interesting in the presence of asymmetries in the face, cases for which X-ray planning performs poorly.

The entry point in the application is the Patient Loader module, in which the surgeon can create and manage patient files and data as it is showed in Figure 3 [Fig. 3] Patient Loader.

Available modules of the application are displayed at the top-left corner of the window. The Osteotomy Planner module is used to perform cutting of bone segments. The 3D Cephalometry module contains the work of this diploma project and is described in higher detail in the next chapters. The Relocation Planner module is used to move bone segments into goal positions.

Below the module selection, a list of patient files is found.

On the right, corresponding studies of the selected patient are listed. A study is a container for patient data referring to a certain planning/working session, i.e. Data Items (planning data), Process Data Items (session data) and Trackable Items (anatomical image/model data).

After clicking on the 3D Cephalometry module activation button and choosing bone segments to be analyzed, the 3D Cephalometry module is presented to the user as seen in Figure 4 [Fig. 4] 3D Cephalometry Module.

A box containing a tab for every type of cephalometric element (e.g. landmarks) is displayed below the module selection area. Each tab shows a list of elements pertaining to the corresponding type that can be shown in the view.

The 3D view occupies most of the screen and displays an interactive perspective projection of the 3D bone segments and cephalometric objects.

The main function in the 3D Cephalometry module is Landmark Definition. Landmarks are the “construction bricks” with which other cephalometric elements are defined (e.g. the distance function is only available if at least two landmarks are defined).

Three landmarks are shown in Figure 5 [Fig. 5] Landmark Definition.

Landmarks are placed by clicking with the mouse onto a bone segment model. They can subsequently be renamed and moved to their optimal position.

A line goes through two landmarks and is defined by clicking on the landmarks. Its appearance in the view is shown in Figure 6 [Fig. 6] Line Definition.

Lines are used as component elements in line-plane angle definitions.

The distance between two landmarks is defined by selecting the landmarks. Its representation in 3D is a double sided arrow with the numerical value floating above it as can be seen in Figure 7 [Fig. 7] Distance Definition.

Distance elements allow the definition of an optimal value range. For example, the surgeon can set the optimal range of a distance to be between 50 and 55mm. If the distance lies inside the optimal range, its color is changed to green, if it lies outside, it is displayed in red. This function is particularly useful when using the Relocation Planner (for an example see pointRelocation Planner).

Angles can be defined in two ways: either by three landmarks or between a line and a plane. Definition by landmarks is shown in Figure 8 [Fig. 8] Angle Definition through Landmarks.

Definiton by angle between line and plane is shown in Figure 9 [Fig. 9] Angle Definition between Line and Plane.

Like distances, angles also allow the definition of an optimal value range. For an example see pointRelocation Planner.

A plane can be defined in two ways: either through 3+n landmarks or with a specific landmark configuration. The first definition supports any number of landmarks thanks to a least-square fitting algorithm. A plane defined by four landmarks can be seen in Figure 10 [Fig. 10] Plane Definition through 3+n Landmarks.

The second available definition mode for planes is a special configuration of five landmarks (first landmark is on the plane, the mean value of the second and third landmark defines the second plane point and the mean of the fourth and fifth landmark for the last point).This configuration is used to define the midsagittal plane as shown in Figure 11 [Fig. 11] Midsagittal Plane Definition through five Landmarks.

The Relocation Planner module is responsible for the planification of the actual bone movements. Saved cephalometry process data can be loaded into this module to provide cephalometric data to the surgeon during planning. The cephalometric elements are constantly and automatically adapted in real-time to the new segment positions in order to reflect the obtained bone profile.

Figure 12 [Fig. 12] Relocation Planner Module with Initial Position shows the Relocation Planner with a set of bone segments located in their initial position. Optimal ranges are defined for the distance and the angle.

When the segment of the jaw is moved, connected cephalometric objects are automatically updated in real-time. As soon as the distance and angle come into the optimal range, the color is changed from red to green indicating, that the segment has now reached an optimal position. Figure 13 [Fig. 13] Relocation Planner Module with Goal Position shows the bone segments after relocation.