Verbessertes Modell eines Neuronalen Graphnetzwerks für Dreidimensionale Dentalmodelle

Details

Veröffentlichungs-Nr.

EP4586219

Aktenzeichen

EP24151356

Anmeldetag

11. Januar 2024

Veröffentlichung

16. Juli 2025

Rechtsraum

EP

IPC

G06V10/82G06N3/045G06V20/64

Abstract

A computer-implemented method for training a graph neural network (GNN) model (106) is provided. The method includes generating a graph (404, 406) corresponding to a three-dimensional (3D) mesh (302) wherein the 3D mesh (302) is associated with a surface information of at least a portion of a dental arch. Further, ground truth data indicating features of the 3D mesh (302) to be predicted by the GNN model (106) is obtained. The graph (404, 406) of the 3D mesh (302) and the ground truth data are provided, as an input, to the GNN model (106) wherein the GNN model (106) comprises a set of convolution layers. Furthermore, a set of weights is applied to each convolution layer of the set of convolution layers, wherein the set of weights corresponds to differences between features of nodes of the graph (404, 406). The method further comprises receiving, as an output of the GNN model (106) with the applied set of weights, output data indicating predicted features of the 3D mesh (302). The method additionally comprises iteratively updating trainable parameters of the GNN model (106) based on comparing the output data to the ground truth data until a predefined loss function threshold is reached.

Anmelder

Firma

3Shape A/S

Land

🇩🇰 Dänemark

Fachgebiete