System for the automatic normalization of dental geometries and extraction of functional occlusal surfaces

Abstract

Computer-implemented data processing system for the automatic normalization of three-dimensional dental geometries, comprising: a) an input interface (110) configured to receive b) a level determination module (120), configured for automatic c) a transformation module (130), configured to transfer the d) a surface filter module (140), configured for selective e) an output interface (150) configured to provide the normalized and filtered dental geometry as a reference model for downstream analysis, simulation or manufacturing systems.

Description

II. Technical Field

[0001] The invention relates to a data processing system and a normalized dental reference model in the field of dental diagnostics and digital dentistry, in particular for computer-aided normalization of intraoral scan data and for the selective extraction of functional occlusal surfaces for downstream analysis and simulation systems. III. State of the art

[0002] The spatial orientation of dental models relative to a reproducible reference plane is a fundamental prerequisite for any functional analysis in dentistry. Conventionally, this orientation is achieved using a facebow, which transfers the spatial position of the maxilla relative to the patient's hinge axis to an articulator.

[0003] Existing Facebow systems (Artex, SAM, Kavo, Denar) require physical patient contact, cost between €200 and €1,500 per device, and require 10–15 minutes of application time per patient. The transmission is prone to positioning errors and requires trained personnel.

[0004] Digital intraoral scanning systems (Medit, iTero, 3Shape, Shining3D, TRIOS) generate three-dimensional surface meshes of the dental arches, but do not provide automatic normalization with respect to a functional reference plane. Furthermore, the scan data contains non-functional areas such as the palate, tooth necks, alveolar ridge, and gingiva, which are irrelevant or disruptive for subsequent functional analyses.

[0005] Existing CAD software (exocad, 3Shape Dental System, CEREC) offers manual tools for axis alignment, but no automatic occlusion plane detection from the tooth geometry itself and no direction-dependent filtering of functional occlusal surfaces.

[0006] Therefore, there is a need for a system that automatically determines the occlusal plane from the existing tooth geometry without external measuring devices, transforms the dental meshes into a standardized coordinate system, and selectively extracts only the functionally relevant occlusal surfaces. IV. Objective of the invention

[0007] The invention is based on the objective of providing a data processing system that automatically determines the occlusal plane from commercially available intraoral scan data, transforms the dental geometries into a normalized coordinate system and extracts functional occlusal surfaces by means of direction-dependent filtering --- without Facebow, without manual alignment and without external measuring devices. V. Solution to the task

[0008] The problem is solved by the data processing system according to claim 1, the normalized reference model according to claim 9 and the arrangement according to claim 12. 1. Determination of occlusion planes (Module 120)

[0009] The system analyzes the tooth geometry of both jaws and automatically determines the occlusal plane. In a preferred embodiment, the contact points between the upper and lower jaws are first identified. Principal component analysis (PCA) of these contact points calculates a best-fit plane, which serves as the occlusal plane. The dental arch geometry is analyzed to determine the sagittal (anterior-posterior), transverse, and vertical axes. An anterior detector identifies the anterior region from the arch shape. 2. Normalization and Coordinate System (Module 130)

[0010] Both jaws are transformed into a common, standardized coordinate system: X-axis transverse (right-left), Y-axis vertical (cranial-caudal), Z-axis sagittal (anterior-posterior). The maxilla establishes the orientation matrix; the mandible adopts this with an inverted Y-axis. Result: The maxilla and mandible share a consistent coordinate system with the occlusal plane as the reference. 3. Direction-dependent surface filtration (Module 140)

[0011] The core innovation: For each surface element (triangle), the surface normal is compared with a separation direction. For maxillary surfaces, it is checked whether the normal points towards the mandible (caudally); for mandibular surfaces, whether it points towards the maxilla (cranially). The filter criterion is the dot product between the surface normal and the separation direction. Only elements with a dot product above a threshold (preferably ≥ 0.2) are classified as functional occlusal surfaces. This filter automatically eliminates: vestibular tooth surfaces, cervical areas, interdental spaces, palate (maxilla), floor of mouth (mandible), and alveolar ridge. The result is pure occlusal surface shells optimized for subsequent functional analyses. 4. Relevance Masking (Module 145)

[0012] A configurable radius of influence around identified contact points or ridge tips determines which areas are considered functionally relevant. The radius is adjustable (preferably 1–15 mm) and defines the export scope. A larger radius (e.g., 8 mm) covers movement amplitudes during protrusion and laterotrusion; a smaller radius (e.g., 3 mm) extracts only the immediate contact areas. 5. Socket generation --- BaseForge (Module 160)

[0013] Optionally, a base generation module creates a closed, watertight reference model. The open occlusal surface shell is supplemented by a standardized digital base: The marginal loop of the occlusal surface is identified (half-edge analysis), the marginal vertices are projected onto a base plane (occlusal plane ± defined distance, preferably 5 mm), a circumferential surface is triangulated between the margin and the projected loop, and a base surface closes the base. The result is a watertight mesh with a defined measurement boundary—the digital equivalent of the standardized plaster base. 6. Browser-based implementation

[0014] A preferred implementation is entirely browser-based. The STL data is processed client-side and does not leave the user's computer (GDPR compliance). Normalization takes place in under one second on standard hardware. VI. List of reference symbols 10 OK-Mesh (upper jaw geometry) 11 UK-Mesh (lower jaw geometry) 100 Data processing system (total system) 110 Input interface 120 Level Determination Module 130 Transformation module 140 Surface filter module 145 Relevance masking module 150 output interface 160 Socket generation module (BaseForge) 170 Export module 20 Normalized Reference Model (Output) 21 Occlusion plane 22 Separation direction (sepDir) 23 Base level 24 radius of influence VII. Brief description of the drawings Fig. 1 --- System block diagram: Data flow through modules 110--170 of the data processing system (100), from the input meshes (10, 11) to the normalized reference model (20). Fig. 2 --- Flowchart: Procedure for determining and normalizing occlusion planes, including PCA, anterior detection and coordinate transformation. Fig. 3 --- Direction-dependent filtering: Schematic representation of the normal filter logic. Left: unfiltered geometry with all surface elements. Right: filtered occlusal surface after application of the scalar product criterion. Arrows indicate accepted (direction of opposing jaw) and rejected (vestibular / lingual) normals. VIII. Summary

[0015] Disclosed are a computer-implemented data processing system (claims 1--8), a normalized dental reference model (claims 9--11) and a browser-based arrangement (claim 12) for the automatic normalization of three-dimensional dental geometries and for the selective extraction of functional occlusal surfaces.

[0016] The system receives three-dimensional dental geometries from commercially available intraoral scanners, automatically determines the occlusal plane by analyzing anatomical structural features, transforms the geometries into a standardized coordinate system, and selectively extracts functional occlusal surfaces using direction-dependent surface filtering. Optionally, a base generation module creates a closed, watertight reference model with a standardized measurement limit. The invention enables the normalization of dental models without a facebow and the generation of analysis-ready reference models from commercially available intraoral scans.

Claims

Computer-implemented data processing system for the automatic normalization of three-dimensional dental geometries, comprising: a) an input interface (110) configured for receiving b) a plane determination module (120) configured for automatic c) a transformation module (130) configured for transferring d) a surface filter module (140) configured for selective e) an output interface (150) configured for providing the normalized and filtered dental geometry as a reference model for downstream analysis, simulation or manufacturing systems. Data processing system according to claim 1, wherein the three-dimensional dental geometry is available as a triangulated surface mesh, as a point cloud, as a volume model or as a combination thereof, in particular in STL, PLY, OBJ or a proprietary scanner format. Data processing system according to claim 1 or 2, wherein the plane determination module (120) determines the occlusal plane by means of at least one of the following methods: principal component analysis (PCA) of the contact points or cusp tips, best-fit plane by means of identified occlusal landmarks, analysis of the dental arch geometry to determine the sagittal and transverse axes, or automatic anterior-posterior recognition from the arch shape. Data processing system according to one of claims 1 to 3, wherein the surface filter module (140) calculates the direction-dependent filter criterion as a scalar product between the surface normal of each surface element and a separation direction, wherein only surface elements are extracted whose scalar product exceeds a predetermined threshold, such that maxillary elements with caudally directed normal and mandibular elements with cranially directed normal are selected. Data processing system according to one of claims 1 to 4, further comprising a relevance masking module (145) configured to determine functionally relevant surface areas by means of a parameterizable influence radius around identified contact points or bump tips, wherein the influence radius is adjustable, in particular, in the range of 1 mm to 15 mm. Data processing system according to one of claims 1 to 5, further comprising a base generation module (160) configured to generate a closed, watertight reference model by: identifying the marginal loop of the filtered occlusal surface, projecting the marginal vertices onto a defined base plane, triangulating a lateral surface between the marginal and the projected loop, and closing a bottom surface, wherein the base plane is defined at a predetermined distance relative to the occlusal plane. Data processing system according to one of claims 1 to 6, wherein the transformation module (130) generates a common orientation matrix for the upper and lower jaw, wherein a first jaw establishes the orientation by means of principal component analysis and anterior detection and the second jaw adopts this orientation with an inverted vertical axis. Data processing system according to one of claims 1 to 7, further comprising an export module (170) configured to output the normalized reference model in at least one standardized three-dimensional data format, in particular as a binary or ASCII-based triangulated surface mesh, wherein the export module applies relevance masking and exports only the surface elements classified as functionally relevant by the surface filter module (140). A normalized dental reference model, stored on a computer-readable medium, comprising: a three-dimensional representation of at least one dental occlusal surface in a normalized coordinate system with respect to an automatically determined occlusal plane, wherein the reference model contains only functionally relevant surface areas selected by direction-dependent filtering, and wherein the reference model is configured as input for a system for determining patient-specific temporomandibular joint parameters, for motion simulation, or for automatic occlusal surface synthesis. Normalized dental reference model according to claim 9, wherein the reference model further comprises a standardized digital base body which complements the reference model to form a closed, watertight mesh, wherein the base body has a defined base plane, a lateral surface and a bottom surface and serves as a reproducible measurement limit for downstream analysis or simulation systems. A normalized dental reference model according to claim 9 or 10, wherein the reference model is configured as input for at least one of the following downstream systems: a system for the inverse determination of condylar parameters (CondylaMap), a system for motion simulation (condylar.pro), a system for automatic occlusal surface synthesis (CrownForge), a system for orthodontic analysis (OrthoGuide), or a system for the fabrication of dental restorations by means of subtractive or additive manufacturing. Arrangement for normalizing dental geometries, comprising a data processing system according to one of claims 1 to 8, wherein the arrangement is executable as a browser-based application on a client computer and all processing steps are performed client-side, so that the dental geometries do not leave the user's computer, or wherein the arrangement is installed as a local application on an end device.

Images