A METHOD FOR A TOOTH RESTORATION
TECHNICAL FIELD
The present invention relates to a method and a system for creating a surface representation for dental applications according to the preamble of claim 1 and 5, respectively.
BACKGROUND
The present invention relates to the field of computer aided design and manufacturing for dental applications. The technique involves scanning a model of an object, e.g. a model of a prepared tooth, transferring the scan data to a computer design system, designing a dental restoration, and transferring design data to a manufac- turing machine for the production of the restoration. CAD/CAM based systems from the design and manufacturing of dental restorations are known in the art, for example:
• Duret: "Vers unit prothese informatisee" Tonus Dentaire No 73, 1985pp. 55-57.
• Duret et al: "CAD-CAM in dentistry", JADA, Vol. 117, November 1988, pp. 715-720.
• Williams: "Dentistry and CAD/CAM: Another French Revolution", Journal of Dental Practice Administration, January/March 1987.
• Sjδlin, Sundh, Bergman: "The Decim System for Production of Dental restorations", International Journal of computerised Dentistry 1999: 3.
In the technical field in question, computer models of dental restorations are often based on the generation of surface representation using a grid system. The grid technology for representing measured topology is well known iii the literature. In "Reverse engineering of geometric models" Varady et al gives an overview of the reverse engineering subject.
As is illustrated in fig. 1 of the enclosed drawings, data in grid format consists of points in a 3D-space (x,y,z), where the x and y points form a regular shaped pattern, e.g. squares with equal sides, while the z co-ordinate is the free variable that de- scribes the topology of the surface.
It is well known to create regular data with grid distribution out of all kinds of irregular data from various types of 3D measuring devices, e.g. contact scanners, optical scanners and others. Such a measuring device is known from WO98/36702. A typical way of achieving this is to calculate, for each grid point, an average coordinate value out of all measurement points in a defined neighbourhood, and assign this value to the grid point.
When this type of surface representation is used in a reverse engineering and CAD/CAM system with the purpose of producing dental restorations, the orientation of the grid is critical, i.e. how the grid co-ordinate system (ex, ey, ez) is oriented with respect to the prepared tooth to be modelled. The direction of the ez -vector is called the insertion axis. If the grid co-ordinate system is not oriented in a good way, parts of the surface may not be possible to represent.
Referring to fig. 2 and 3 in the enclosed drawings, the following example will illustrate this problem. Fig. 2 and 3 shows schematically a cross section of a tooth 1 prepared for an inlay, the tooth presenting a cavity 2. The part of the tooth that is possible to model is shown with a thicker line. With the orientation of the grid co- ordinate system shown in fig. 2, the complete prepared cavity can be modelled.
Fig. 3 shows the same cross section as in fig. 2, but here the grid is oriented in a slightly different way. The effect of this is that a part of the preparation, illustrated by dotted lines, can not be modelled. Here, this part is called an undercut region with respect to the insertion axis, i.e. the direction of the ez -vector. Thus, to achieve
good results in a CAD/CAM-system, using a grid for the surface representation, it is essential that the insertion axis is chosen correctly.
In a reverse engineering and CAD/CAM system for the dental application, the grid is generated out of the measurement data. Often, the grid orientation is hard coded with respect to the scanner co-ordinate system. This implies that the direction of the grid must be determined at the time of scanning. Thus, the insertion in the scanning apparatus of the object to be scanned becomes a critical task that takes time and requires special training.
SUMMARY OF THE INVENTION
The object of the present invention is to reduce processing time and requirements on special training for staff involved when creating a surface representation of an ob- ject for dental applications.
This object is obtained by a method and a system presenting the characterizing features of claims 1 and 5, respectively.
The invention makes it possible to find a grid orientation which minimizes or eliminates, for a given set of scan data, the undercut regions, and generates a grid that handles the remaining undercut regions in a way that facilitates the manufacturing of a restoration with a good fit.
Moreover, this can be done also in the case of the insertion axis orientation, at the scanning stage being poorly chosen in relation to the scanned object. This means that the scanning does not have to be repeated in such a case. Thus, as a result of the use of the invention, the operator in the scanning procedure is released from the burden of choosing a grid orientation. This will turn the scanning into a simple step that takes shorter time than when using conventional technique. Also, the scanning
may be done by personnel without a lot of dental training and experience, or it can easily be automated.
BRIEF DESCRIPTION OF THE FIGURES
The invention will now be described in detail with the aid of the enclosed drawings, in which:
- fig. 1 shows schematically a 3 -dimensional grid system,
- fig. 2 shows schematically a cross section of a prepared tooth, - fig. 3 shows schematically a cross section of a prepared tooth,
- fig. 4 shows schematically an arrangement for collecting topographical data about an object,
- fig. 5 shows schematically an object about which topographical data is collected,
- fig. 6 shows a part of the object in fig. 5 in a cross section view, - fig. 7 shows schematically a computer model of the object in fig. 5, and
- fig. 8 shows a part of the object in fig. 5 in a cross section view.
DETAILED DESCRIPTION
Fig. 11 shows a schematic view of a known arrangement to collect topographic data of a plaster model of a tooth prepared at a dentist, the model here being referred to as a preparation 3. A laser emitter 4 is located above the preparation 3. The laser emitter 4 projects laser radiation in a plane P, and a camera 5 can focus on a line K defined by the intersection of the surface of the preparation 3 and the plane P. Dur- ing the scanning procedure the preparation 3 is moved in relation to the plane P, in the normal direction of the plane P itself.
After the plane has been moved over the entire preparation 3, the preparation is either rotated around an axis parallel to the z-axis in fig. 1, or tilted around an axis being parallel to the x-y-plane, as described in the Patent Application SE0003575-8.
After the rotation or tilting operation the preparation 3 is scanned again. The scanning is repeated a number of times with intermediate rotations or tilting of the object.
An object of repeating the scanning procedure of the preparation 3 is to cover the entire surface of the object in the collecting of topographical data. That is, parts of the surface not covered in one scanning procedure, e.g. because they are hidden due to the angle of the camera in relation to the object, can be covered in another scanning procedure after rotation or tilting of the object. Accordingly, a plurality of "overlapping" sets of scan data is obtained, resulting in the obtaining of topographical information over the entire surface of the scanned object.
Alternatively the collecting of topographical data can be done, using another type of optical method, such as white light projection with fringes. Also a non-optical method, such as touch probe scanning, can be used,
Fig. 5 shows a model of a preparation 3, presenting a surface irregularity in the form of a cavity 6 on one of its sides. Fig. 6 shows a part of a section of the preparation 3, the section being oriented along the line L in fig. 5. Also, in fig. 6 a cross section of the cavity 6 is shown. Preferably, in a computer modelling stage, a first surface representation is created, using the scan data. The surface representation is created using a grid with a three-dimensional grid co-ordinate system (x, y, z). The orientation of the grid can be randomly chosen. Preferably, a first grid orientation is chosen for the first surface representation, corresponding to the orientation of a scan co- ordinate system used in the scan data, i.e. according to the position of the preparation 3 in the reader.
As an example which can be seen in fig. 6, for a value of the x and y co-ordinate, xl and yl, there are three z- values, zl, z2 and z3 corresponding to the surface of the preparation 1. Preferably, the highest z-value is chosen for the first surface repre-
sentation, in this case z3. Later there is a need for identification of the part or parts of the first surface representation, where a z- value has been chosen in said manner. Therefore, preferably, all points where the highest z-value has been chosen in said manner are marked, or stored, to be able to discern these points in subsequent steps of the method. Continuous regions of points with z- values chosen in this manner, are here referred to as undercut areas U. Since only the highest z-value is chosen for the first surface representation, the latter will not show the cavity 6, rather it will present a discontinuity, represented by the line D in fig. 6, in the region of the cavity 6.
Fig. 31 shows the computer model of the preparation, generates with the first surface representation. It can clearly be seen that the first grid orientation causes a "drop" in the surface in the region of the cavity 6. In one alternative of the inventive method the computer operator detects the "drop" and concludes that an undercut re- gion is present there. Also, undercut regions, tagged in the manner described above, can be detected by the software marking them on the display, e.g by indicating them with a separate color. The operator can then generate a second surface representation after entering commands for changing the orientation of the grid, to decrease or eliminate to undercut area at the cavity 6. The detection of undercut areas could also be done by the computer software used to generate the surface representation, in a manner described above.
In the event of more than one undercut area in the first surface representation, all of them are identified. Preferably, a main undercut area MU is identified. Preferably, the main undercut area is determined to be the undercut area having the largest surface area out of all undercut areas U. Alternatively, the main undercut area could be the undercut area having the largest projected surface onto the x-y-plane, or the one having the greatest length in the direction of its largest extension. Here we assume that fig. 6 shows a main undercut area MU.
A new grid orientation is chosen for a second surface representation. The aim is to eliminate or decrease the presence of multiple z- values corresponding to one x-y- value. To establish a direction in which to turn the grid, at least one z-value zL of the first surface representation, on one side of the discontinuity D, or the main un- dercut area MU is selected. This z-value zL is compared to at least one other z-value zH of the first surface representation on the opposite side of the discontinuity D, or the main undercut area MU. The location of the part of the first surface representation with the lower z-value zL indicates the direction, in which to turn the grid. The axis around which to turn the grid can be determined as an axis parallel to the direc- tion of the largest extension of the main undercut area MU.
Referring to fig. 8, the grid is turned in the direction established previously, by an angular distance, which can be predetermined, or registered or stored from a user input. A second surface representation is generated. If undercut areas present them- selves on the second surface representation, the method has to be executed again, until a surface representation with no, or minimal, undercut areas is obtained.
In some cases it might be impossible to avoid undercut areas totally, and the aim of the method is then to minimise these. At a dental restoration manufactured in such a case, the spaces between the restoration and the prepared tooth at the undercut areas, would be filled with cement.