Method for analyzing an image of a dental arch

A method for evaluating the suitability of a substantially transparent dental aligner comprising:

• A) instructing the taking of a plurality of images of at least a portion of a first person's dental arch with a mobile device associated with the first person:
• A′) wherein the taking of the plurality of images is assisted by a first application on the mobile device that identifies a suitable acquisition condition;
• A″) wherein the taking of the plurality of images is at least partially performed while the first person is using a dental retractor;
• A′″) wherein at least one image of the plurality of images is taken with the substantially transparent dental aligner in a service position;
• A″″) wherein the taking of the plurality of images does not occur at a dental office;
• B) receiving by a second application an output relating to an analysis of at least one image of the plurality of images of at least a portion of the first person's dental arch from a deep learning device capable of evaluating the suitability of the substantially transparent dental aligner, wherein the analysis comprises:
• B′) identifying at least one separation between the substantially transparent dental aligner and the first person's dental arch;
• wherein the identification of the at least one separation between the substantially transparent dental aligner and the first person's dental arch is first analyzed without using contour-based image processing;
• B″) relating the at least one separation identified in the at least one image of the plurality of images to a physical measurement by using information known independently of the at least one image of the plurality of images;
• B′″) determining information about the fit of the substantially transparent dental aligner at least in part based on the at least one separation between one or more teeth in the first person's dental arch and the substantially transparent dental aligner in the service position;
• B″″) wherein the analysis is performed by a deep learning device trained on a set of at least ten thousand historical dental images each having labelled separations between a dental arch and a substantially transparent dental aligner in the service position;
• and
• C) displaying at the mobile device a message related to the analysis;
• C′) wherein the displaying at the mobile device the message related to the analysis is performed without action from a dental professional;
• C″) wherein the message displayed at the mobile device relates to whether the substantially transparent dental aligner is suitable to the current orthodontic treatment.

A method for training a deep learning device to identify a separation between a substantially transparent dental aligner in a service position and at least a portion of a dental arch comprising:

• A) collecting a learning base of at least ten thousand images comprising at least one attribute;
• A′) wherein in at least a subset of the at least ten thousand images, the at least one attribute includes a separation between a substantially transparent dental aligner and one or more teeth in a dental arch;
• A″) wherein the at least one attribute is identified by a label in the subset of the at least ten thousand images;
• B) providing the learning base to the deep learning device to train the deep learning device, wherein the deep learning device is a neural network;
• C) receiving at least one additional image with a label associated with at least one attribute;
• C′) wherein the at least one additional attribute associated the additional image includes a separation between the substantially transparent dental aligner and one or more teeth in a first person's dental arch;
• D) sending updates to the learning base that include information from the additional image;
• and
• E) using the deep learning device that has been updated based on updates to the learning base to identify a separation between a substantially transparent dental aligner in the service position and at least a portion of a dental arch;
• F) displaying at the mobile device a message related to the analysis.