Method and system of teeth alignment based on simulating of crown and root movement
The present disclosure relates to a dental image processing protocol for the design of dental aligners. Specifically, the dental image processing protocol aids in the determination of tooth movements during realignment, based on an initial position and a final position, and on characteristics of the periodontal environment. Therefore, planned tooth movements reflect both crown movement and root movement within biological structures of the alveolar process.
1. A method for generating an intermediate position of one or more teeth of a dental arch of a patient, comprising:
classifying, via processing circuitry, pixels of one or more medical images into classes corresponding to biological structure types;
segmenting, via the processing circuitry, the classified pixels of the one or more medical images into biological structures;
rendering, via the processing circuitry, a three-dimensional model of the biological structures based on the segmented classified pixels;
determining, via the processing circuitry, one or more metrics, based upon the three-dimensional model, describing characteristics of the biological structures;
acquiring, via the processing circuitry, a final position of each of the one or more teeth of the dental arch of the patient based upon the three-dimensional model; and
generating, via the processing circuitry, the intermediate position of each of the one or more teeth of the dental arch of the patient based upon the one or more metrics and the acquired final position,
wherein one of the one or more metrics is
a distance metric defined as a distance between a surface of a root of one of the one or more teeth of the dental arch of the patient and a surface of a neighboring biological structure of the three-dimensional model, or
a density metric defined as a measure of mean intensity of voxels adjacent to a central voxel.
2. The method according to claim 1 , wherein the classifying further comprises
training, via the processing circuitry, a classifier on a training database; and
classifying, via the processing circuitry, the pixels of the one or more medical images based upon the classifier,
wherein the training database includes a corpus of reference medical images, each reference medical image comprising at least one identifiable reference biological structure associated in the training database with at least one corresponding description of a biological structure type.
3. The method according to claim 2 , wherein the training further comprises
training, via the processing circuitry, a first neural network according to a first dataset;
training, via the processing circuitry, a second neural network according to a second dataset, the second dataset comprising a plurality of classification predictions of the first neural network; and
generating, via the processing circuitry, the training database based upon a plurality of classification predictions of the second neural network.
4. The method according to claim 3 , wherein the second neural network is a fully convolutional neural network.
5. The method according to claim 1 , wherein the intermediate position of each of the one or more teeth of the dental arch of the patient is determined based upon a position of an aspect of a neighboring tooth of the one or more teeth of the dental arch of the patient.
6. The method according to claim 1 , wherein the surface of the neighboring biological structure of the three-dimensional model is a surface of an alveolar process.
7. An apparatus for processing of dental images, comprising:
processing circuitry configured to
classify pixels of one or more medical images into classes corresponding to biological structure types;
segment the classified pixels of the one or more medical images into biological structures;
render a three-dimensional model of the biological structures based on the segmented classified pixels;
determine one or more metrics, based upon the three-dimensional model, describing characteristics of the biological structures;
acquire a final position of each of one or more teeth of a dental arch of a patient based upon the three-dimensional model; and
generate an intermediate position of each of the one or more teeth of the dental arch of the patient based upon the one or more metrics and the acquired final position,
wherein one of the one or more metrics is
a distance metric defined as a distance between a surface of a root of one of the one or more teeth of the dental arch of the patient and a surface of a neighboring biological structure of the three-dimensional model, or
a density metric defined as a measure of mean intensity of voxels adjacent to a central voxel.
8. The apparatus according to claim 7 , wherein the processing circuitry is further configured to
train a classifier on a training database; and
classify the pixels of the one or more medical images based upon the classifier,
wherein the training database includes a corpus of reference medical images, each reference medical image comprising at least one identifiable reference biological structure associated in the training database with at least one corresponding description of a biological structure type.
9. The apparatus according to claim 8 , wherein the training further comprises
training a first neural network according to a first dataset;
training a second neural network according to a second dataset, the second dataset comprising a plurality of classification predictions of the first neural network; and
generating the training database based upon a plurality of classification predictions of the second neural network.
10. The apparatus according to claim 7 , wherein the intermediate position of each of the one or more teeth of the dental arch of the patient is determined based upon a position of an aspect of a neighboring tooth of the one or more teeth of the dental arch of the patient.
11. The apparatus according to claim 7 , wherein the surface of the neighboring biological structure of the three-dimensional model is a surface of an alveolar process.