Three-dimensional dental model segmentation quality assessment
AssigneeAlign Technology, Inc.
InventorsAndrei SEMENOV, Denis DURDIN, Jason RAMOS, Sergio SANTAMARIA VENEGAS
Life of this patent
Estimating the term…
Abstract
Provided herein are apparatuses (e.g., systems) and methods for automatically assessing a segmented 3D dental model. The segmented 3D dental model may be classified by a trained machine learning classifier (a trained neural network). In some cases, the segmented 3D dental model may be classified as a passing (acceptable) or failing (unacceptable) model. The machine learning classifier may be trained with labeled 3D dental models that illustrate passing and failing segmented 3D dental models.
Claims27
A method, the method comprising:
- receiving a segmented three-dimensional (3D) dental model of a subject's dentition;
- classifying the segmented 3D dental model using one or more trained machine learning classifier engines, wherein the one or more machine learning classifier engines are trained on one or more of:
- trimming plane errors, scan artifact, extra teeth, and missing teeth;
- and
- passing the segmented 3D dental model to a treatment planning engine based on the classification, otherwise re-segmenting the 3D dental model.
A method, the method comprising:
- receiving a segmented three-dimensional (3D) dental model of a subject's dentition, a scan data of the subject's dentition, and an age classification of the subject;
- classifying the segmented 3D dental model using one or more trained machine learning classifier engines, wherein the one or more machine learning classifier engines are trained using a training database comprising automatically segmented 3D dental models, corresponding scan data, and age classification scored for and one or more of:
- missing or incorrectly placed trimming planes, excess or absence of non-dental anatomic material, extra teeth, and missing teeth;
- and
- passing the segmented 3D model to a treatment planning engine based on the classification, otherwise re-segmenting the 3D dental model.
A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processor to perform a method comprising:
- receiving a segmented three-dimensional (3D) dental model of a subject's dentition;
- classifying the segmented 3D dental model using one or more trained machine learning classifier engines, wherein the one or more machine learning classifier engines are trained on one or more of:
- trimming plane errors, scan artifact, extra teeth, and missing teeth;
- and
- passing the segmented 3D model to a treatment planning engine based on the classification, otherwise re-segmenting the 3D dental model.
A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processor to perform a method comprising:
- receiving a segmented three-dimensional (3D) dental model of a subject's dentition, a scan data of the subject's dentition, and an age classification of the subject;
- classifying the segmented 3D dental model using one or more trained machine learning classifier engines, wherein the one or more machine learning classifier engines are trained using a training database comprising automatically segmented 3D dental models, corresponding scan data, and age classification scored for and one or more of:
- missing or incorrectly placed trimming planes, excess or absence of non-dental anatomic material, extra teeth, and missing teeth;
- and
- passing the segmented 3D model to a treatment planning engine based on the classification, otherwise re-segmenting the 3D dental model.
Dependencies are inferred from the claim text.