DE10225854A1 - Method for production of medical prostheses in which a model is based on medical image data and an iterative process is followed to ensure a 3-D model for use with a CNC machine accurately matches the input medical data - Google Patents

Abstract

Method for construction of a medical prosthesis in which medical image data is converted to a 3-D output model that is converted into a matching prosthesis model via an intermediate step so that a final prosthesis can be produced using a CNC machine. The intermediate step comprises a number of partial steps in which a first 3-D comparison model is compared with actual reference points on the 3-D medical output model. The comparison model is then adjusted until it is nearer to the medical output model. The steps are interated until agreement between the models is reached. The resulting model is then used by the CNC machine.

Description

The invention relates to a method for the rapid construction of medical prostheses and manufactured by this method Prostheses.

In human medicine there is often the task of accident or Disease damaged or lost body parts, tissues or to replace other functional structures, such as bones.

Here it is necessary to use the parts that represent the original structures replace, to design under different criteria.

The main aspects are the restoration and the Maintaining the original functionality as well as often cosmetic rehabilitation.

In the cases mentioned, the task is to be as precise as possible Perform the shape of the prosthesis. This step should be from an objective point of view, as automatically as possible, with a reasonable effort, a few steps and with inclusion the anatomical present on the individual patient Shape relationships.

It is known that, based on medical data such as CT and MRI, a prosthesis specially developed for a patient getting produced.

In US 5,741,215 a method is described in which on the Basis of a patient's medical image data via a stereolithographic model a corresponding prosthesis or anatomical model can be created.

This document also mentions a method in which, in the course of Determining the shape of the prosthesis to correct an existing one Defect the symmetrical side opposite the defect Attention.  

It is also known that when determining the shape of such Prostheses can also be used for other comparison models can. This is especially necessary if, due to the location of the defects to be closed symmetry properties of human Bodies can not be used.

In US 5,824,085 a method is described in which the defect the implant to be inserted is adjusted. One becomes Database of standard parts (hip joints) an approximate one Implant selected and according to its shape a model and the location of the matching counterpart determined and made by a robot.

In a similar way, the one disclosed in DE 199 22 279 A1 uses Process a database of comparison models, in which the most similar with the help of a few typical anatomical points of an algorithm for the search for similar parts is preselected.

In US 5,798,924 a method is mentioned in which starting of medical and anatomical data, models of the defect area and the desired result can be adjusted against each other. This is only due to the approximation of both models area data oriented at support points.

All of the aforementioned methods have in common that various complex manufacturing steps are necessary to produce a corresponding prosthesis ( FIG. 1).

With increasing number of such steps the temporal as well equipment expenditure until the completion of the to be used Product. The number of used for the implant construction Support points are z. T. drastically reduced, although the Output data provide considerably more information.

Each additional step in the process chain also increases the total error. Steps, such as the generation of the 3D model, the interactive reconstruction of the data and generation of the actual implant data additionally require interactive intervention of the human.  

The invention has for its object a method for rapid Construction of medical prostheses to indicate that starting medical image data, fast, accurate and automatic Shape design for prostheses enables.

According to the invention, this object is achieved by the characterizing Features of the first claim solved. advantageous Refinements are made by subordinate claim 2 specified.

The essence of the invention is that by the automatic Sequence of individual, explained in more detail below, successive procedural steps the interaction (the intervention of humans) is kept to a minimum during implant production is taking into account all information of the output data is guaranteed and the generation of the virtual shape of the prosthesis in can be carried out within a few hours.

As with the known methods, the starting point is specific medical image data of the patient. These are, for example, shift data computer tomography. These data are in a subsequent Step to a three-dimensional volume model of the patient put together, all for the generation of the prosthesis model contains the necessary data. This includes in particular bone structure as well as the reconstruction of the soft tissue (e.g. muscles, nerves, Tendons, blood vessels, etc.).

At this point, the method according to the invention starts by the initial model, which is based on specific medical Image data from patients is determined in a known manner in 3-D, and a first 3-D comparison model any reference point in local similarly located regions of both models and the distance between the two reference points and the directional orientation of both reference points to each other is determined. Based on the determined The shift of these two reference points to each other becomes the Shift of the data points of the 3-D output model to the first 3-D Transfer D comparison model.  

These steps of the method according to the invention are used for generation of n 3-D comparison models repeated n times, the so generated n 3-D comparison models used to generate the prosthesis model by making a differentiated comparison of the n 3-D comparison models to each other to form a single model, which in a known manner is used for the CNC-supported production of the prosthesis.

The advantage of the method according to the invention is that the use of a complete volume data set ensures that all accessible information of patient data for prosthesis production can be used because there is no data reduction (such as with Polygon models or surface models) is carried out.

The invention is described below with reference to the drawings and the Exemplary embodiments explained in more detail. Show it:

Fig. 1 sequence of known methods for the construction of medical prostheses,

Figure 2 is a general history. Of the inventive method,

Fig. 3 process steps of determining geometry,

FIG. 4a producing a 3-D model and comparison

FIG. 4b generating the prosthesis model from two 3-D models comparison.

FIG. 5a skull defect: output model,

Fig. 5b skull defect: comparison model

Fig. 6 of error in the vicinity of the skull defect,

Fig. 7 prosthesis model.

In the process, specific medical image data from Patients converted into a 3-D output model in a known manner. On this converted 3-D output model and a first 3-D  Comparative model is a reference point arbitrarily in locally similar regions of both models.

Then the next step is the distance between the two Reference points and the directional orientation of both reference points determined to each other. The displacement vectors are by amount given as well as direction and can be determined by determining the specific Volume focal points are determined, always a clear assignable volume centers can be found.

Based on the determined shift of these two reference points to one another, the shift of the data points of the 3-D output model is transferred to the first 3-D comparison model ( FIG. 4a).

These partial steps of the method can be repeated n times to generate n 3-D comparison models, the n 3-D comparison models generated in this way, as shown by way of example in FIG. 4b for n = 2, being used to generate the prosthesis model by A differentiated comparison of the n 3-D comparison models to one another takes place to a single model.

This differentiated comparison is made by comparing the n 3-D Comparative models to the 3-D starting model carried out, whereby spatial data are determined in their entirety the CNC trimmed production of a prosthesis.

The data acquisition for the manufacture of a prosthesis takes place in the following steps:

• - Definition of specific points in the initial and comparison model (e.g. with easily identifiable points of the models, see g. "Landmarks");
• - Determination of the shift (identified by amount and Direction) of the points in the initial and comparison model;
• - Application of the "inverse" shift to neighboring points in the Comparison model and thereby calculation of points for the Comparison model;
• - Determination of the deviation between the comparison model and Initial model and determination of the deviation error;  
• - comparison of the deviation error with a predeterminable value, Repeat all previous steps if the Error deviation is too large;
• - If the error is minimized, the determined points are on all Points of the comparison model transformed;
• - Creation of the prosthesis model by Boolean subtraction from Initial model and comparison models.

The peculiarity of the process is:

• - Depending on the selection of the starting points, different "contours" or comparison models result. An automatic correction of the edge points and a scaling of the entire point set is necessary ( FIG. 4a);
• - Determination of a finite number n of comparison models (because of 3D maximum number = number of edge points);
• - Determination and assignment of "closely" neighboring points in the Compare models;
• - Calculation of an error measure (deviation, distances) between the individual comparison models (e.g. sum of the quadratic Deviation);
• - The most important criterion is "constant transition" of the connection to the Edge points Introduction of a weighting of the influence of the points each by distance from the edge when determining the error measure (Edge points weighting = 1 with distance from edge weighting - <0);
• - Specification of a minimum deviation to be achieved;
• - Determination of the shifts between the n boundary points and off the end point of the "opposite" edge point generated comparison model;
• - Applying a weighted shift to points of the comparison model which decreases with distance from the edge point (in Fig. 4b to about half of the curve of the edge model weighting one, from then zero).

LIST OF REFERENCE NUMBERS

1

Initial model of the skull

2

Initial model of the bone defect

3

Comparative model of the bone defect

4

prosthesis model

Claims (2)

1. A method for the rapid construction of medical prostheses, in which specific medical image data of patients are converted into a 3-D output model and an appropriate prosthesis model is generated via an intermediate step, on the basis of which a prosthesis is produced by means of CNC-assisted production, characterized in that that the sub-steps a) to d) are carried out as follows in the intermediate step:

• a) on the starting model and a first 3-D comparison model, a reference point is defined in regions of the same location in both models;
• b) the distance between the two reference points and the directional orientation of the two reference points is determined;
• c) on the basis of the determined displacement of these two reference points relative to one another, the displacement of the data points of the 3-D output model is transferred to the first 3-D comparison model;
• d) the sub-steps a) to c) are repeated n times to generate n 3-D comparison models, the n 3-D comparison models thus generated being used to generate the prosthesis model by performing a differentiated comparison of the n 3-D - Comparative models to each other are made into a single model.

2. The method according to claim 1, characterized in that the differentiated comparison by comparing the n 3-D Comparison models to the 3-D output model is carried out, whereby spatial data are determined which in their entirety the CNC-assisted manufacturing of a prosthesis.