阅读说明：本技术 牙齿类型和萌出状态的自动检测 (Automatic detection of tooth type and eruption status ) 是由 薛亚 崔智英 J·B·摩尔 A·斯彼德诺夫 于 2018-06-15 设计创作，主要内容包括：本文提供了用于检测目标牙齿的萌出状态(例如,牙齿类型和/或萌出状态)的系统和方法。可以对患者的牙齿进行扫描和/或分割。可以识别目标牙齿。可以提取牙齿特征、主成分分析(PCA)特征和/或其他特征,并将其与其他牙齿的上述特征(例如通过自动机器学习系统获得的那些特征)进行比较。检测器可以识别和/或输出目标牙齿的萌出状态,例如目标牙齿是完全萌出的乳牙、部分萌出/未萌出的恒牙还是完全萌出的恒牙。(Systems and methods for detecting an eruption state (e.g., tooth type and/or eruption state) of a target tooth are provided herein. The patient's teeth may be scanned and/or segmented. The target tooth may be identified. Tooth features, Principal Component Analysis (PCA) features, and/or other features may be extracted and compared to the above-described features of other teeth (e.g., those obtained by an automated machine learning system). The detector may identify and/or output the eruption status of the target tooth, e.g. whether the target tooth is a fully erupted deciduous tooth, a partially erupted/unerupted permanent tooth or a fully erupted permanent tooth.)

1. A computer-implemented method, comprising:

collecting a 3D model of a patient's dentition;

identifying a portion of the 3D model of the patient's dentition corresponding to the target tooth, the portion of the 3D model of the patient's dentition being associated with one or more visual attributes of the target tooth;

normalizing the 3D model of the target tooth;

identifying one or more principal component analysis features of a portion of the 3D model of the patient's dentition corresponding to the target tooth, the one or more principal component analysis features being related to the one or more visual properties of the target tooth;

determining one or more tooth eruption indicators for the target tooth using the one or more principal component analysis features, the one or more tooth eruption indicators providing a basis for identifying an eruption state of the target tooth; and

outputting the one or more tooth eruption indicators.

2. The computer-implemented method of claim 1, wherein normalizing comprises identifying a predetermined number of angles relative to a center point of the target tooth.

3. The computer-implemented method of claim 1, wherein outputting the one or more tooth eruption indicators comprises outputting an indicator corresponding to one of: eruption of deciduous teeth, partial eruption or non-eruption of permanent teeth, and eruption of permanent teeth.

4. The computer-implemented method of claim 1, wherein collecting 3D models comprises one or more of: obtaining a 3D model of a patient's teeth; receiving a 3D model of a patient's teeth from an intraoral scanner; and receiving the 3D model from a scan of the patient's dental model.

5. The computer-implemented method of claim 1, wherein the eruption state comprises one or more of an eruption state and a persistence type of a target tooth.

6. The computer-implemented method of claim 5, wherein the persistent state specifies whether the target tooth is a permanent tooth or a deciduous tooth.

7. The computer-implemented method of claim 1, wherein determining the one or more tooth eruption indicators for a target tooth is based at least in part on patient age, eruption sequence, measured space available for eruption, patient gender, or other patient information related to the patient.

8. The computer-implemented method of claim 1, wherein determining the one or more tooth eruption indicators comprises comparing the one or more principal component analysis features to one or more principal component analysis features of a 3D model of one or more representative teeth.

9. The computer-implemented method of claim 1, wherein determining the one or more tooth eruption indicators for a target tooth based on principal component analysis features comprises applying a machine learning algorithm selected from the group consisting of decision trees, random forests, logistic regression, support vector machines, AdaBOOST, K-nearest neighbors, quadratic discriminant analysis, and neural networks.

10. The computer-implemented method of claim 1, wherein determining the one or more tooth eruption indicators for target teeth comprises using a machine-trained classifier to identify one or more of an eruption state and a tooth type of target teeth based on one or more principal component analysis features of a 3D model of one or more representative teeth.

11. The computer-implemented method of claim 10, wherein the machine-trained classifier implements a convolutional neural network.

12. The computer-implemented method of claim 1, further comprising outputting a modified version of the 3D model of the patient's teeth to include tooth numbers based at least in part on the eruption state of the target teeth.

13. The computer-implemented method of claim 1, further comprising receiving a request to identify the one or more tooth eruption indicators for a target tooth; and wherein identifying one or more principal components is performed in response to the request.

14. The computer-implemented method of claim 1, further comprising using the one or more tooth eruption indicators as part of an orthodontic treatment plan for a patient's teeth.

15. The computer-implemented method of claim 14, wherein the orthodontic treatment plan comprises a pediatric orthodontic treatment plan for a pediatric patient.

16. The computer-implemented method of claim 1, further comprising using the one or more tooth eruption indicators to design at least a portion of a removable orthodontic appliance for a patient's dentition.

17. The computer-implemented method of claim 1, wherein outputting the one or more tooth eruption indicators comprises one or more tooth eruption indicator tags corresponding to the one or more tooth eruption indicators.

18. The computer-implemented method of claim 1, wherein identifying the portion of the 3D model is part of an operation of segmenting the 3D model of the patient's dentition.

19. A system, comprising:

one or more processors;

a memory coupled to the one or more processors, the memory configured to store computer program instructions that, when executed by the one or more processors, implement a computer-implemented method comprising:

collecting a 3D model of a patient's dentition;

identifying a portion of the 3D model of the patient's dentition corresponding to the target tooth, the portion of the 3D model of the patient's dentition being associated with one or more visual attributes of the target tooth;

identifying one or more principal component analysis features of a portion of the 3D model of the patient's dentition corresponding to the target tooth, the one or more principal component analysis features being related to the one or more visual properties of the target tooth;

determining one or more tooth eruption indicators for the target tooth using the one or more principal component analysis features, the one or more tooth eruption indicators providing a basis for identifying an eruption state of the target tooth; and

outputting the one or more tooth eruption indicators.

20. A method of automatically determining the eruption state and deciduous or permanent tooth type of a target tooth, the method comprising:

receiving, in a computing device, a 3D model of a patient's teeth including a target tooth;

determining, in a computing device, tooth shape features of a target tooth from a 3D model of a patient's tooth;

determining, in a computing device, tooth shape features of one or more reference teeth from a 3D model of a patient's teeth;

normalizing, in the computing device, at least some of the tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth;

applying the normalized tooth shape features to a classifier of a computing device; and

the eruption state of the target tooth and the deciduous or permanent tooth type are output from the computing device.

21. The method of claim 20, wherein outputting comprises outputting one of: eruption of deciduous teeth, partial eruption or non-eruption of permanent teeth, and eruption of permanent teeth.

22. The method of claim 20, further comprising obtaining a 3D model of the patient's teeth.

23. The method of claim 20, wherein the 3D model is received from a three-dimensional scanner.

24. The method of claim 20, wherein the 3D model is received from a mold of a patient's teeth.

25. The method of claim 20, further comprising obtaining patient information, wherein the patient information includes one or more of: patient age, emergence sequence, measured space available for emergence, and patient gender; wherein applying the normalized tooth shape features to the classifier comprises applying the patient information to the classifier using the normalized tooth shape features.

26. The method of claim 20, wherein determining the tooth shape characteristic of the target tooth comprises determining one or more of: mesio-distal width, facial-lingual width, crown height, crown center, and number of cusps.

27. The method of claim 26, wherein determining a number of cusps comprises determining a number of cusps in one or more arch directional surfaces comprising: buccal mesial, buccal distal, lingual mesial, and lingual distal.

28. The method of claim 20, wherein determining the tooth shape characteristics of one or more reference teeth comprises determining the tooth shape characteristics of one reference tooth.

29. The method of claim 20, wherein the one or more reference teeth comprise molars.

30. The method of claim 20, wherein determining the tooth shape characteristics of one or more reference teeth comprises determining the tooth shape characteristics of two reference teeth.

31. The method of claim 20, wherein determining the tooth shape characteristics of the one or more reference teeth comprises: for each of the one or more reference teeth, determining one or more of: mesio-distal width, facial-lingual width, and crown center.

32. The method of claim 20, wherein normalizing at least some of the tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth comprises normalizing one or more of the mesial-distal width, the facial-lingual width, the crown height, and the crown center to the one or more reference teeth.

33. The method of claim 20, wherein normalizing further comprises determining a total number of cusps in each arch direction surface, the arch direction surfaces comprising: buccal mesial, buccal distal, lingual mesial, and lingual distal.

34. The method of claim 20, wherein applying the normalized tooth shape features to a classifier comprises applying a first level binary classifier or a first level binary classifier and a second level binary classifier to the normalized tooth shape features.

35. The method of claim 20, wherein applying the normalized tooth shape features to the classifier comprises: applying a first level binary classifier to the first subset of normalized tooth shape features and indicating an eruption state and a deciduous or permanent tooth type of the target tooth based on the first level binary classifier, or applying a second level binary classifier to the second subset of normalized tooth shape features and indicating an eruption state and a deciduous or permanent tooth type of the target tooth based on the second level binary classifier.

36. The method of claim 20, wherein outputting an eruption status comprises outputting an indication of a percentage of eruption.

37. A method of automatically determining the eruption state and deciduous or permanent tooth type of a target tooth, the method comprising:

receiving, in a computing device, a 3D model of a patient's teeth including a target tooth;

determining, in a computing device, tooth shape features of a target tooth from a 3D model of a patient's tooth;

determining, in a computing device, tooth shape features of one or more reference teeth from a 3D model of a patient's teeth;

normalizing, in a computing device, at least some of the tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth;

applying the normalized tooth shape features to a classifier of a computing device, wherein applying the normalized tooth shape features to the classifier comprises: applying a first level binary classifier and outputting the eruption state and the deciduous or permanent tooth type of the target tooth according to an output of the first level binary classifier, or applying a second level binary classifier to the normalized tooth shape feature and then outputting the eruption state and the deciduous or permanent tooth type of the target tooth.

38. A method of automatically determining the eruption state and deciduous or permanent tooth type of a target tooth, the method comprising:

receiving a three-dimensional model of a patient's teeth including a target tooth;

normalizing at least one dimension of the target tooth based on the one or more reference teeth;

inputting the tooth shape features including the at least one dimension to a first binary classifier to determine whether the target tooth is a fully erupted permanent tooth;

inputting tooth shape features into a second binary classifier to determine whether the target tooth is a partially erupted/unerupted permanent tooth or a deciduous tooth if the first binary classifier determines that the target tooth is not a fully erupted permanent tooth; and

the output target teeth are fully erupted permanent teeth, partially erupted/unerupted permanent teeth, or deciduous teeth.

39. A non-transitory computing device readable medium having instructions stored thereon for determining a state of a target tooth of a patient, wherein the instructions are executable by a processor to cause a computing device to:

receiving a 3D model of a patient's teeth including a target tooth;

determining tooth shape characteristics of the target tooth from the 3D model of the patient's tooth;

determining tooth shape features of one or more reference teeth from the 3D model of the patient's teeth;

normalizing at least some tooth shape features of a target tooth using tooth shape features of the one or more reference teeth;

applying the normalized tooth shape features to a classifier of a computing device; and

outputting the eruption state of the target tooth and the deciduous or permanent tooth type.

40. The non-transitory computing device-readable medium of claim 39, wherein the instructions are further configured to cause the output to be one of: deciduous teeth erupting, permanent teeth partially erupting/not erupting, or permanent teeth erupting.

41. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to receive a 3D model from a three-dimensional scanner.

42. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to obtain patient information, wherein the patient information comprises one or more of: patient age, emergence sequence, measured space available for emergence, and patient gender; wherein the instructions are further configured to include in the patient information the normalized tooth shape features applied to the classifier.

43. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to determine, as part of the tooth shape feature, one or more of: mesio-distal width, facial-lingual width, crown height, crown center, and number of cusps.

44. The non-transitory computing device-readable medium of claim 39, the instructions further configured to determine a number of cusps by determining a number of cusps in one or more arch directional surfaces, the arch directional surfaces comprising: buccal mesial, buccal distal, lingual mesial, and lingual distal.

45. The non-transitory computing device readable medium of claim 39, wherein the instructions are configured to determine tooth shape features of the one or more reference teeth from one reference tooth.

46. The non-transitory computing device readable medium of claim 45, wherein the one reference tooth comprises a molar.

47. The non-transitory computing device readable medium of claim 39, wherein the instructions are configured to determine the tooth shape features of the one or more reference teeth from two reference teeth.

48. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to determine tooth shape features of one or more reference teeth for one or more of: mesio-distal width, facial-lingual width, and crown center.

49. The non-transitory computing device-readable medium of claim 39, wherein the instructions are further configured to: at least some of the tooth shape features of the target tooth are normalized using the tooth shape features of the one or more reference teeth by normalizing one or more of the mesio-distal width, facial-lingual width, crown height, and crown center to the one or more reference teeth.

50. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to normalize at least some tooth shape features of a target tooth by determining a total number of cusps in each arch-oriented surface, the arch-oriented surfaces comprising: buccal mesial, buccal distal, lingual mesial, and lingual distal.

51. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to apply the normalized tooth shape features to a classifier by applying a first level binary classifier or a first level binary classifier and a second level binary classifier to the normalized tooth shape features.

52. The non-transitory computing device readable medium of claim 39, wherein the instructions are further configured to apply a first level binary classifier to the first subset of normalized tooth shape features and indicate the eruption state of the target tooth and the deciduous or permanent tooth type based on the first level binary classifier, or apply a second level binary classifier to the second subset of normalized tooth shape features and indicate the eruption state of the target tooth and the deciduous or permanent tooth type based on the second level binary classifier.

53. The non-transitory computing device-readable medium of claim 39, wherein the instructions are further configured to output an indication of a percentage of eruption.

Background

In pediatric cases, orthodontic appliances are often used before all of the patient's permanent teeth erupt. Detecting erupting teeth may help prevent the appliance from striking or contacting the erupting teeth, which may stop eruption of teeth or harm the patient.

In many prior art techniques, medical professionals manually determine the eruption state of a patient's teeth and the type of teeth. This may require physical examination of the teeth or images/scans of the teeth and medical professionals to make judgments about eruption status and tooth type, which adds time and expense.

Disclosure of Invention

Drawings

The novel features believed characteristic of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present disclosure will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the disclosure are utilized, and the accompanying drawings of which:

fig. 1A is a diagram illustrating an example of a computing environment configured to digitally scan an arch and determine a tooth type and/or eruption state of a target tooth.

Fig. 1B is a diagram illustrating an example of a segmentation engine.

Fig. 1C is a diagram showing an example of a feature extraction engine.

Fig. 1D is a diagram illustrating another example of a feature extraction engine.

Fig. 1E is a diagram showing an example of a detector engine.

Fig. 1F is an example of a method of automatically determining the eruption state and/or tooth type of a target tooth.

Fig. 2A is an example of a method of automatically determining the eruption state and tooth type of a target tooth.

Fig. 2B is an example of a method of automatically determining the eruption state and tooth type of a target tooth.

Fig. 2C-2D are flow charts describing detectors that use first and second binary classifiers to determine the eruption state and tooth type of a target tooth. The detector receives the normalized features of the target tooth to determine the eruption state and the tooth type.

FIG. 3A is a flow chart describing a method of extracting and generating data to be passed into a tooth state classification detector.

Fig. 3B is a flow chart describing a tooth condition classification detector that determines the eruption condition and tooth type of a target tooth using data from the flow chart of fig. 3A.

Fig. 3C is a flowchart describing a classifier of the tooth condition classification detector that determines the eruption state and the tooth type of the target tooth.

Fig. 4 shows the distribution of mesial-distal widths of the teeth 4.

Fig. 5A shows a distribution of crown heights of the tooth 6.

Fig. 5B shows the distribution of the bucco-lingual width of the tooth 6.

Fig. 6 shows the distribution of the crown height of the tooth 5.

Fig. 7 shows examples of tooth types and tooth shapes in the erupted state, including 1) permanent teeth, full eruption, 2) permanent teeth, partial eruption, and 3) deciduous teeth.

FIG. 8 is a simplified block diagram of a data processing system that may perform the methods described herein.

Embodiments address the need to provide an automated tooth detection eruption detection system that can automatically, efficiently, and accurately predict tooth types and eruptions of dental patients with high accuracy. The present application addresses these and other technical problems by providing a solution and/or an automated agent for automatically detecting the tooth type and eruption status of a dental patient. Automatic detection of tooth condition (e.g., tooth type and/or eruption condition) may provide a basis for implementing an automatic orthodontic treatment plan, designing and/or manufacturing orthodontic appliances, including a series of polymeric orthodontic appliances that provide a force to correct malocclusions of a patient's teeth.

In general, the example apparatuses (e.g., devices, systems, etc.) and/or methods described herein may receive representations of a patient's teeth and, in some cases, clinical information about the patient to determine one or more indicators of tooth eruption associated with the patient's teeth. As used herein, a "tooth eruption indicator" may include an indicator of teeth associated with eruption. Examples of the tooth eruption indicator include an eruption state of teeth. Examples of eruption states include eruption states (e.g., whether teeth have fully erupted, partially erupted, or not erupted at all) and eruption persistence states ((e.g., whether teeth are deciduous/young or permanent). Including eruption and deciduous/permanent tooth types.

As used herein, a "patient" can be any subject (e.g., human, non-human, adult, child, etc.), and can alternatively and equivalently be referred to herein as a "patient" or a "subject. As used herein, a "patient" may be, but is not necessarily, a medical patient. As used herein, a "patient" may include a person receiving orthodontic treatment, including orthodontic treatment with a series of orthodontic appliances.

Any of the devices and/or methods described herein may be part of a distal tooth scanning device or method, or may be configured for use with a digital scanning device or method.

As will be described in greater detail herein, automatically determining the eruption state of the target teeth (e.g., for each patient's teeth) and/or tooth types such as deciduous or permanent teeth may include collecting a 3D scan of the patient's teeth. Collecting the 3D scan may include performing a 3D scan, including scanning the arch of the patient's teeth directly (e.g., using an intraoral scanner) or indirectly (e.g., scanning an impression of the patient's teeth), receiving 3D scan information from a separate device and/or a third party, receiving a 3D scan from memory, and/or the like.

Additional information, including patient information (e.g., age, gender, etc.), may be collected using the 3D scan.

The 3D scan information may be normalized (normalized) or normalized (normalized). Normalizing the scan may include converting the 3D scan to a standard format (e.g., a tooth surface mesh), and/or representing the 3D scan as a plurality of angles (e.g., vector angles) from a center point of each tooth, and/or the like. In some variations, normalizing may include normalizing the 3D scan using another tooth, including the stored tooth values.

The normalized 3D scan information may then be processed to extract one or more features that may be used to automatically determine tooth eruption, partial eruption or non-eruption, and/or tooth type; specifically, the tooth is a deciduous tooth or a permanent tooth. This information can be used to automatically and accurately mark the teeth of the 3D model, for example, by numbering the teeth with standard tooth numbers.

For example, a method of automatically determining an eruption state and a deciduous or permanent tooth type of a target tooth may include: collecting, in a computing device, a three-dimensional (3D) model of a patient's teeth including a target tooth; normalizing the 3D model of the target tooth; applying Principal Component Analysis (PCA) of the dental features of the target tooth in the computing device to obtain PCA features; determining the eruption state and deciduous or permanent tooth type of the target tooth based on the PCA characteristics; and outputting the eruption state of the target tooth and the deciduous or permanent tooth type.

Note that although examples using PCA are provided herein, other feature vector based multivariate analysis may also be used. PCA is proposed because it can be performed automatically using known techniques including the use of correlation and/or covariance techniques, iterative methods including, but not limited to, nonlinear iterative partial least squares techniques, to compute PCA.

Normalization may include identifying a predetermined number of angles relative to a center point of the target tooth. The center of the tooth may be determined using any suitable method. For example, the center of a tooth may be determined from a representation of the mesh points from each tooth (e.g., from a segmented version of a 3D scan of a digital model representing the patient's teeth), by determining the geometric center of the mesh points for each tooth, by determining the center of gravity of the segmented tooth, and so forth. The same method for determining the center of each tooth can be applied consistently to the teeth as well as to any teeth used to form (e.g., train) the system described herein.

Normalization may be different from normalization. As used herein, normalization may involve adjusting the number and/or other description of the teeth. For example, normalization may involve adjusting the number of orders and/or angles (from the center of the tooth) used to describe the teeth. The tooth size from the original 3D scan can be preserved.

Any of the methods and systems described herein can determine whether each tooth in the arch is a deciduous or permanent tooth and/or whether the tooth erupts, does not erupt, or partially erupts. Thus, for one or more (e.g., all) of the patient's teeth in a 3D scan of the patient's dental arch, any of these methods may determine that the tooth is a deciduous erupting, a permanent partial eruption/non-eruption, and/or a permanent eruption tooth. In any of these methods and systems, for each tooth for which eruption and/or deciduous/permanent tooth descriptions have been determined, the system may use this information to mark a digital model of the patient's teeth with an index. For example, the apparatus and/or methods described herein may output (via display, transmission, etc.) a 3D model of a patient's teeth, labeling each tooth as one of: a primary tooth eruption (or "primary tooth only"), a permanent tooth partial eruption/non-eruption, and/or a permanent tooth eruption (or "permanent tooth only").

The 3D scan of the patient's teeth may be collected in any suitable manner that allows a later method or apparatus operating it to perform normalization, feature extraction, and determination of eruption and/or permanent/deciduous tooth status. For example, collecting may include obtaining a 3D model of the patient's teeth directly or indirectly from the patient's teeth. For example, the collecting may include receiving a 3D model of the patient's teeth from an intraoral scanner. The collecting may include receiving a 3D model from a mold scan of the patient's teeth.

Any of the devices and/or methods described herein may include collecting patient information about a patient, wherein the patient information includes one or more of: patient age, emergence sequence, measured space available for emergence, and patient gender; in addition, wherein determining the eruption state of the target tooth and the tooth type such as deciduous or permanent teeth is further based on the patient information. Patient information may be obtained directly or indirectly from the patient, including querying and/or receiving patient questionnaire information, extracting patient information for electronic patient records, and the like.

Any suitable features may be extracted from the prepared (e.g., normalized and/or normalized) teeth. For example, in some variations, the features may include a Principal Component Analysis (PCA) for each tooth in the arch being examined. Additional features (e.g., geometric descriptions of the patient's teeth) may not be needed (e.g., PCA alone may be used), or additional features may be used to supplement the PCA of each tooth. As described above, PCA may be automatically performed on standardized teeth using any suitable technique, including using modules from existing software environments such as C + + and C # (e.g., ALGLIB libraries implementing PCA and truncated PCA, MLPACK)), Java (e.g., KNIME, Weka, etc.), Matmatica, MATLAB (e.g., MATLAB statistics toolkit, etc.), python (e.g., Matplotlib python library, Scikit-left, etc.), GNU Octave, etc.

In any of the devices and methods described herein, eruption status and deciduous or permanent tooth types may be automatically determined (using one or more processors) based on information extracted from the standardized 3D scan. The 3D model of the patient's teeth may be segmented and/or normalized prior to or as part of the normalization process. For example, a 3D scan of a patient's dental arch may be transmitted to and processed by a segmentation engine that will divide the 3D scan into individual teeth and/or gums.

For example, the method and/or apparatus may determine the eruption state and deciduous or permanent tooth type of the target tooth based on PCA features by applying a machine learning algorithm selected from the group consisting of decision trees, random forests, logistic regression, support vector machines, AdaBOOST, K-nearest neighbors (KNNs), quadratic discriminant analysis, and neural networks. For example, machine learning techniques can be used to form and apply trained networks (neural networks). Alternatively or additionally, logistic regression may be used.

A system (e.g., a system for determining eruption status and tooth type such as deciduous or permanent teeth) may include: one or more processors; a memory coupled to the one or more processors, the memory configured to store computer program instructions that, when executed by the one or more processors, implement a computer-implemented method comprising: collecting a three-dimensional (3D) model of a patient's teeth including a target tooth; normalizing the 3D model of the target tooth; performing Principal Component Analysis (PCA) on the dental features of the target tooth to obtain PCA features; determining the eruption state and deciduous or permanent tooth type of the target tooth based on the PCA characteristics; and outputting the eruption state of the target tooth and the deciduous or permanent tooth type. Any of these systems may include a memory for storing results (e.g., 3D models of labeled teeth). Any of these systems may also include output devices (e.g., monitors, printers, transmitters, including wireless transmitters), and so forth.

The devices and/or methods described herein may also be performed using one or more sets of reference teeth for normalization. For example, the methods described herein may include methods of automatically determining the eruption state and deciduous or permanent tooth type of a target tooth. The method can comprise the following steps: receiving a three-dimensional (3D) model of a patient's teeth including a target tooth in a computing device, determining tooth shape features of the target tooth in the computing device from the 3D model of the patient's teeth, determining tooth shape features of one or more reference teeth in the computing device from the 3D model of the patient's teeth, normalizing at least some of the tooth shape features of the target tooth in the computing device using the tooth shape features of the one or more reference teeth, applying the normalized tooth shape features to a classifier of the computing device, and outputting an eruption state of the target tooth and a deciduous or permanent tooth type from the computing device.

In any of the devices and/or methods described herein, the automatic determination of eruption state, particularly in the steps of receiving, determining tooth shape characteristics, normalizing, etc., may be performed using a device (e.g., a computing device) without manual control or guidance. Alternatively or additionally, any of these steps may be performed partially automatically (e.g., semi-automatically) or manually.

The computing device may receive, directly (e.g., as part of a scanning device or system) or indirectly (including transfer from previously obtained models), a three-dimensional model (3D) of the patient's teeth including the target teeth. The computing device may be a dedicated device or part of a dedicated device (e.g., a scanner), or may be connected, either wired or wirelessly, to the scanning device or to a memory that stores the scanning information. Alternatively or additionally, the computing device may receive the 3D model from a remote (e.g., internet, cloud, etc.) source.

In any of the devices and/or methods described herein, the target tooth may be selected by a user. Alternatively or additionally, all teeth in the 3D model of teeth may be selected as targets; the apparatus and method may determine the eruption state and deciduous or permanent tooth state of each of a plurality of target teeth (including all teeth) sequentially or simultaneously.

In any of the apparatuses and/or methods described herein, an output may be provided that includes outputting one of: deciduous teeth eruption, partial eruption of permanent teeth (including permanent teeth not erupting, collectively referred to herein as partial eruption/non-eruption of permanent teeth), and permanent teeth eruption. The output may include manufacturing an orthodontic and/or dental appliance based on the determined tooth condition (e.g., primary tooth eruption, partial eruption or non-eruption of permanent teeth, and permanent tooth eruption). The predicted spacing may be performed based on a determined eruption of deciduous teeth, partial eruption or non-eruption of permanent teeth, and eruption of permanent teeth. The spacing may be determined based on existing and neighboring tooth models (3D models) and/or empirical tooth information from a particular patient or group of similar patients. The eruption information and/or spacing may be used to determine a model (3D model, including digital and non-digital models, such as a physical model) that may be manufactured and/or used in the manufacture of the teeth and/or orthodontic devices.

The methods described herein may also include obtaining a 3D model of the patient's teeth. A 3D model may be received from a three-dimensional scanner. Alternatively or additionally, the 3D model may be received from a mold of the patient's teeth.

The methods described herein may include obtaining patient information, wherein the patient information includes one or more of: patient age, emergence sequence, measured space available for emergence and patient gender; further wherein applying the normalized tooth shape features to the classifier comprises applying the patient information to the classifier using the normalized tooth shape features.

Determining the tooth shape characteristic of the target tooth may include determining one or more of: mesio-distal width, facial-lingual width, crown height, crown center, and number of cusps. Determining the number of cusps may include determining the number of cusps in one or more arch-oriented surfaces, the arch-oriented surfaces including: buccal mesial, buccal distal, lingual mesial, and lingual distal.

Determining the tooth shape characteristics of one or more reference teeth may include determining the tooth shape characteristics of one reference tooth. The one or more reference teeth may comprise molars. Determining the tooth shape features of the one or more reference teeth may include determining the tooth shape features of two reference teeth. Any suitable tooth shape characteristic (morphological characteristic) may be determined. For example, determining the tooth shape features of the one or more reference teeth may include determining, for each of the one or more reference teeth, one or more of: mesio-distal width, facial-lingual width, and crown center.

Normalizing at least some of the tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth may include normalizing one or more of the mesial-distal width, the facial-lingual width, the crown height, and the crown center to the one or more reference teeth. The normalization may also include determining a total number of cusps in each arch direction surface, the arch direction surfaces including: buccal mesial, buccal distal, lingual mesial, and lingual distal.

Any of these methods may further include applying the normalized tooth shape features to a classifier, including: applying the first-level binary classifier or the first-level binary classifier and the second-level binary classifier to the normalized tooth shape features. Applying the normalized tooth shape features to the classifier may include: applying a first level binary classifier to the first subset of normalized tooth shape features and indicating an eruption state and a deciduous or permanent tooth type of the target tooth based on the first level binary classifier, or applying a second level binary classifier to the second subset of normalized tooth shape features and indicating an eruption state and a deciduous or permanent tooth type of the target tooth based on the second level binary classifier.

Any of these methods may further include outputting the eruption status, which includes outputting an indication of a percentage of eruption.

Any method of automatically determining the eruption state and deciduous or permanent tooth type of a target tooth may include: receiving in a computing device a three-dimensional (3D) model of a patient's teeth including a target tooth, determining in the computing device tooth shape features of the target tooth from the 3D model of the patient's teeth, determining in the computing device tooth shape features of one or more reference teeth from the 3D model of the patient's teeth, normalizing in the computing device at least some of the tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth, applying the normalized tooth shape features to a classifier of the computing device, wherein applying the normalized tooth shape features to the classifier comprises applying a first-level binary classifier, and outputting an eruption state of the target tooth and a deciduous or permanent tooth type depending on an output of the first-level binary classifier, or applying a second-level binary classifier to the normalized tooth shape features, then outputting the eruption state of the target tooth and the deciduous or permanent tooth type.

A method of automatically determining an eruption state and a deciduous or permanent tooth type of a target tooth may include: the method includes receiving a three-dimensional model of a patient's teeth including a target tooth, normalizing at least one dimension of the target tooth based on one or more reference teeth, inputting tooth shape features including the at least one dimension to a first binary classifier to determine whether the target tooth is a fully erupted permanent tooth, inputting the tooth shape features to a second binary classifier to determine whether the target tooth is a permanent tooth that is a partially erupted/unerupted permanent tooth or a deciduous tooth if the first binary classifier determines that the target tooth is not a fully erupted permanent tooth, and outputting whether the target tooth is a fully erupted permanent tooth, a partially erupted/unerupted permanent tooth, or a deciduous tooth.

Also described herein is an apparatus (including software, firmware, and/or hardware configured to control the apparatus to perform and/or coordinate the performance of any of the methods described herein) for performing the methods. For example, described herein are non-transitory computing device readable media having instructions stored thereon for determining a target dental state for a patient. The instructions are executable by the processor to cause a computing device to receive a three-dimensional (3D) model of a patient's tooth including a target tooth, determine tooth shape features of the target tooth from the 3D model of the patient's tooth, determine tooth shape features of one or more reference teeth from the 3D model of the patient's tooth, normalize at least some tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth, apply the normalized tooth shape features to a classifier of the computing device; and outputting the eruption state of the target tooth and the deciduous or permanent tooth type.

The instructions may also be configured such that the output is one of: deciduous teeth erupting, permanent teeth partially erupting/not erupting, or permanent teeth erupting.

The instructions may also be configured to receive a 3D model from a three-dimensional scanner. The instructions may be configured to obtain patient information, wherein the patient information includes one or more of: patient age, emergence sequence, measured space available for emergence, and patient gender; and further wherein the instructions are configured to include the patient information and the normalized tooth shape features applied to the classifier.

The instructions may also be configured to determine, as part of the tooth shape feature, one or more of: mesio-distal width, facial-lingual width, crown height, crown center, and number of cusps.

In any of these devices, the instructions may be further configured to determine the number of cusps by determining a number of cusps in one or more arch-oriented surfaces, the arch-oriented surfaces comprising: buccal mesial, buccal distal, lingual mesial, and lingual distal.

The instructions may be configured to determine tooth shape features of one or more reference teeth from one reference tooth. One reference tooth may comprise a molar.

The instructions may be configured to determine tooth shape features of one or more reference teeth from two reference teeth.

The instructions may also be configured to determine tooth shape features of the one or more reference teeth for one or more of: mesio-distal width, facial-lingual width, and crown center.

The instructions may also be configured to normalize at least some tooth shape features of the target tooth using the tooth shape features of the one or more reference teeth by normalizing one or more of the mesio-distal width, facial-lingual width, crown height, and crown center to the one or more reference teeth.

The instructions may also be configured to normalize at least some tooth shape features of the target tooth by determining a total number of cusps in each arch-oriented surface, the arch-oriented surfaces including: buccal mesial, buccal distal, lingual mesial, and lingual distal.

The instructions may also be configured to apply the normalized tooth shape features to the classifier by applying the first-level binary classifier or the first-level binary classifier and the second-level binary classifier to the normalized tooth shape features.

The instructions may also be configured to apply a first level binary classifier to the first subset of normalized tooth shape features and indicate the eruption state and the deciduous or permanent tooth type of the target tooth based on the first level binary classifier, or apply a second level binary classifier to the second subset of normalized tooth shape features and indicate the eruption state and the deciduous or permanent tooth type of the target tooth based on the second level binary classifier.

In any of the apparatuses described herein, the instructions may be further configured to output an indication of the percentage of eruption. In general, the output may be visual and/or digital and/or printed.