Method and system for aligning digital 3D models of teeth

Abstract

A computer-implemented method for aligning digital three- dimensional models of a patient's dentition is disclosed. The method comprises receiving a first digital three-dimensional model (101) of the patient's dentition comprising a first plurality of tooth regions and receiving a second digital three- dimensional model (102) of the patient's dentition comprising a second plurality of tooth regions. The method further comprises obtaining closest tooth regions by identifying, for each tooth region of the first plurality of tooth regions, a closest tooth region of the second plurality of tooth regions. Further, the method comprises identifying corresponding tooth regions (200) of the first digital three-dimensional model (101) and the second digital three-dimensional model (102) by identifying, from the closest tooth regions, those closest tooth regions with a relative deviation in a surface feature lower than a threshold for the surface feature, wherein the surface feature is a surface normal and wherein the threshold is a threshold for surface normal selected in a range of 5 degrees to 30 degrees. Additionally, the method comprises aligning the first digital three-dimensional model (101) and the second digital three- dimensional model (102) by minimizing distances between the identified corresponding tooth regions (200), and displaying the aligned first digital three-dimensional model (101) and the second digital three-dimensional model (102).

Description

[0001] METHOD AND SYSTEM FOR ALIGNING DIGITAL 3D MODELS OF TEETH

[0002] Technical field

[0003] The disclosure relates to a computer-implemented method and system for aligning digital three-dimensional models of a patient' s teeth.

[0004] Background

[0005] In the field of digital dentistry, use of three-dimensional scanners such as intraoral scanners, is essential for producing digital representations of patient's teeth. These digital representations are digital three-dimensional models of the dental situation, including teeth and gingiva. Based on the digital representations of teeth, dental professionals can perform a wide range of dental treatments accurately and effectively. An important aspect of processing the digital three-dimensional models of teeth is the capability of aligning two or more of the digital three-dimensional models representing teeth of the same patient scanned at different times.

[0006] Alignment of at least two digital three-dimensional models of teeth of the same patient, in a common three-dimensional space, is desired as it enables dental practitioners to observe dental conditions such as tooth wear, dental plaque, gingivitis, dental recession, tooth cracks and / or caries. Additionally, dental practitioners can track the progress of an orthodontic treatment in which teeth are moved towards a desired position.

[0007] Rigidly aligning two three-dimensional models is a challenging task. Known alignment methods are based on identifying corresponding points on the digital three-dimensional models utilizing a distance metric and minimizing distances between such points. The resulting alignment may however be inaccurate because some of the corresponding points, identified solely based on the distance metric, are points with non-matching geometry. Change in geometry may occur due to anatomical changes on the teeth, for example caused by tooth wear.

[0008] Known alignment algorithms are prone to converging to local minima and depend heavily on initial guesses. For this reason, improved methods of aligning the digital three-dimensional models of teeth are desired.

[0009] Summary

[0010] In an embodiment, a computer-implemented method for aligning digital three-dimensional models of a patient's dentition is disclosed, the method comprising:

[0011] - receiving a first digital three-dimensional model of the patient's dentition comprising a first plurality of tooth regions ,

[0012] - receiving a second digital three-dimensional model of the patient's dentition comprising a second plurality of tooth regions ,

[0013] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model based on comparing a surface feature of the first plurality of tooth regions with the surface feature of the second plurality of tooth regions,

[0014] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0015] Term "digital three-dimensional model of a patient's dentition" refers to a digital, three-dimensional (3D) , computer-generated representation of the patient's dental situation including teeth and gingiva.

[0016] Such digital 3D models may accurately correspond to the patient's actual dentition so that dental objects like teeth, teeth surfaces, restorations and gingiva on the digital 3D models correspond to those of the actual dentition at the time of scanning.

[0017] A digital 3D model may be constructed by the processor, based on scan data collected in an intraoral scanning process in which an intraoral 3D scanner may be used to scan the patient's dental situation comprising teeth and gingiva. The intraoral 3D scanner throughout the disclosure is also referred to as the intraoral scanner. The digital 3D model can be stored in a memory of a computer system, for example in a Standard Triangle Language (STL) format.

[0018] The digital 3D model can be received or accessed by the processor. The digital 3D model may usually be displayed on a display screen in form of a 3D mesh, representing surfaces of teeth and gingival tissue. The 3D mesh may be comprised of individual facets, for example triangular facets while each facet comprises, for example, three mutually connected vertices. Alternatively, the digital 3D model may be displayed as a point cloud comprising points, a graph comprising nodes and edges, a volumetric representation comprising voxels, or any other suitable 3D representation form. The digital 3D model may be converted from one representation into another, for example from the 3D mesh into the point cloud by collapsing vertices of the 3D mesh.

[0019] In the context of the disclosure, the first and second digital three-dimensional models of the patient's dentition represent the dentition of the same patient but obtained by scanning the patient' s dentition at different times, for example in an interval of six months or a year. A digital three-dimensional model of the dentition may also be referred to as a digital representation of the patient's dentition. The patient's dentition may comprise teeth of at least lower or upper patient' s jaw.

[0020] The method of the disclosure may comprise obtaining the first digital three-dimensional model of the patient's dentition by scanning the patient' s dentition with an intraoral scanner at a first time and obtaining the second digital three-dimensional model of the patient's dentition by scanning the patient's dentition with the intraoral scanner at a second time, later than the first time.

[0021] The first digital three-dimensional model and the second digital three-dimensional model may be visualized in a common three- dimensional space.

[0022] Corresponding tooth regions of the first digital three- dimensional model and the second digital three-dimensional model, in the context of the disclosure, may be understood as tooth regions of corresponding teeth with matching geometry or most similar geometry. Corresponding teeth are teeth with the same Universal Numbering System (UNN) notation of the first digital three-dimensional model and the second digital three- dimensional model. Thus, corresponding tooth regions relate to geometrically matching tooth regions of the corresponding teeth. For example, the corresponding tooth regions may represent those parts of the patient's dentition with relatively unchanged anatomy during the period in-between obtaining the first digital three-dimensional model and the second digital three-dimensional model. These corresponding tooth regions may be used for accurately aligning the first digital three-dimensional model and the second digital three-dimensional model.

[0023] The corresponding tooth regions may be identified by comparing one or more local surface features of the first digital three- dimensional model and the second digital three-dimensional model. Based on the comparison of one or more local surface features, local variation in geometry of the first digital three-dimensional model and the second digital three-dimensional model may be detected.

[0024] A tooth region may relate to a point or a plurality of points on the surface of a tooth of a digital three-dimensional model. The tooth region may also relate to a tooth facet or a plurality of tooth facets of a tooth of the digital three-dimensional model. In general, the tooth region may be a building element of the digital three-dimensional model such as a point or a facet, or a plurality of connected (neighboring) building elements.

[0025] A surface feature of a tooth region, also related as a surface parameter, may be any one of : a surface normal representing local orientation of the tooth region, a surface curvature of the tooth region, a surface roughness of the tooth region, a color of the tooth region or spin images of the tooth region.

[0026] The surface curvature may measure a local change of direction of the tooth region and may be useful for identifying a peak, a pit, a saddle, a ridge and / or a flat part of the tooth region. The surface roughness usually measures a variation in height over the tooth region. The color of the tooth region may be obtained from color information comprised in the first and / or the second digital three-dimensional representation. Spin images may relate to two-dimensional images of neighboring tooth regions viewed from a local perspective.

[0027] Comparing the surface feature of the first plurality of tooth regions with the surface feature of the second plurality of tooth regions may comprise identifying, for each tooth region of the first plurality of tooth regions, a tooth region of the second plurality of tooth regions with a minimal deviation in the surface feature.

[0028] Deviation in the surface features relates to the relative difference in value of the surface feature of the two tooth regions being compared. For example, a tooth region from the first plurality of tooth regions may be compared to every tooth region from the second plurality of tooth regions in order to find a unique tooth region from the second plurality of tooth regions having the minimal deviation in the surface feature with respect to the tooth region from the first plurality of tooth regions. The same process may be repeated for the rest of tooth regions of the first plurality of tooth regions. In this way it is possible to identify tooth regions of the first and second digital three-dimensional model with the most similar geometry where similarity in geometry is quantified through the relative difference in value of the surface feature. Subsequent alignment of the digital three-dimensional models may be based on aligning the thereby-identified corresponding tooth regions.

[0029] Thus, identifying corresponding tooth regions may comprise comparing the surface feature of each tooth region of the first plurality of tooth regions to the surface feature of each tooth region of the second plurality of tooth regions.

[0030] In an example, the computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0031] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0032] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0033] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, for each tooth region of the first plurality of tooth regions, the tooth region of the second plurality of tooth regions with the minimal deviation in surface normal ,

[0034] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0035] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0036] Deviation in surface normal relates to relative difference in value of the surface normal of the two tooth regions being compared .

[0037] In a further example, the computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising: - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0038] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0039] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, for each tooth region of the first plurality of tooth regions, the tooth region of the second plurality of tooth regions with the minimal deviation in surface curvature ,

[0040] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions,

[0041] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0042] Deviation in surface curvature relates to relative difference in value of the surface curvature of the two tooth regions being compared .

[0043] In a further embodiment, a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0044] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0045] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions , identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions to obtain closest tooth regions,

[0046] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, from the closest tooth regions, those closest tooth regions with a relative deviation in the surface feature lower than the threshold for the surface feature,

[0047] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions,

[0048] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0049] The identified closest tooth regions may be referred to as presumed corresponding tooth regions. The method may thus comprise identifying presumed corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model .

[0050] Closest tooth regions may refer to those tooth regions on the first digital three-dimensional model and the second digital three-dimensional model with shortest distances between them. Identifying closest tooth regions on the digital three- dimensional models may therefore be based on applying a distance criterion, in which shortest distances between the digital three-dimensional models are found. These presumed corresponding tooth regions may or may not be the corresponding tooth regions with matching geometry in the context of the disclosure, to be used in the subsequent alignment step. That is because the distance criterion alone is not sufficiently reliable in identifying geometrically corresponding tooth regions on the first and second digital three-dimensional model. A further step may therefore be required, to filter out, from the identified presumed corresponding tooth regions, the actual corresponding tooth regions .

[0051] The filtering of presumed corresponding tooth regions may be based on one or more surface features. Filtering may be performed by identifying those closest tooth regions with a relative deviation in the surface feature lower than the threshold for that surface feature.

[0052] The method may thus further comprise filtering, from the plurality of presumed corresponding tooth regions, the corresponding tooth regions by identifying those presumed corresponding tooth regions with a relative deviation in the surface feature lower than the threshold for the surface feature. Filtering may, in this case, be referred to as correspondence filtering and may be based on one or more of the surface features.

[0053] The threshold for the surface feature may relate to any one of: a threshold for surface normal, a threshold for surface curvature, a threshold for surface roughness, a threshold for color, a threshold for spin images. In general, the threshold for the surface feature may be a tunable parameter derived from previous cases of well-aligned digital representations.

[0054] To accurately align the first digital three-dimensional model and the second digital three-dimensional model, some relative deviation in the observed surface features may be tolerated as such small relative deviation may be related to inaccuracy of scanners used to scan the dentition.

[0055] However, if the relative deviation in the surface feature between presumed corresponding tooth regions is above the threshold for that surface feature, then it may be concluded that significant change in geometry between the presumed corresponding tooth regions being compared exists, and that those tooth regions with the significant change in geometry should not be used for the purpose of aligning the first digital three-dimensional model and the second digital three-dimensional model. For example, the patient may lose a part of their lower- left canine due to tooth wear, which may be observed on the second digital three-dimensional model and not on the first digital three-dimensional model. Then, only tooth regions on the digital representations of that canine tooth which are unaffected by tooth wear should be used to align the first and second digital representations of that canine tooth. The regions affected by the tooth wear, which may be detected by comparing corresponding surface features, should not be considered in the alignment .

[0056] Correspondence filtering may be applied to the presumed corresponding tooth regions to identify, among the presumed corresponding tooth regions, the actual corresponding tooth regions characterized by relative deviation in the surface curvature below the threshold for surface curvature. This threshold for surface curvature may be set to a value of 0.3 1 / mm for mean curvature. Another example of correspondence filtering may be identifying, from the presumed corresponding tooth regions, those tooth regions with relative deviation in the surface normal below the threshold for surface normal, for example below 30 °. These presented threshold values are shown to be a good compromise between the noise imposed by scanning process and surface linearization, and the deviation that indicates a change of the real-world geometry. In general, the threshold values can be tuned depending on the scanner and software used.

[0057] Ideally, the threshold for surface normal could be chosen such that there is very littly or no deviation in the surface normals when identifying corresponding tooth regions. For example, the threshold could be set to a value up to 5 degrees (°) . However, the alignment algorithm requires sufficient resistance to the noise introduced by the intraoral 3D scanner during the scanning process. This means that, in reality, very few or no corresponding tooth regions would be identified with such value of the threshold.

[0058] In this embodiment, the presumed corresponding tooth regions satisfying the defined threshold may be simply referred to as the corresponding tooth regions or active corresponding tooth regions, while the presumed corresponding tooth regions falling outside the defined threshold may be referred to as non-active correspondences and may be discarded from the alignment process.

[0059] In an embodiment, the threshold for surface normal may be variable. For example, the value of the threshold for surface normal may be lowered as the iterations of the alignment algorithm progress. For example, the threshold for the surface normal may linearly decrease with progressing iterations of the alignment algorithm. Any other function of threshold decrease may be set, such as non-linear function, step function etc.

[0060] It may be noted that for further improvement in algorithm behavior, the correspondence filtering may be based on more than one surface feature. An example would be to use both the surface normal and the surface curvature as surface features, with corresponding thresholds mentioned above. Therefore, a matching condition may be defined if relative deviation in every considered surface feature is below the corresponding threshold.

[0061] If relative deviation in at least one surface feature is above the threshold for that surface feature, the matching condition is not met. For example, the corresponding tooth regions may be identified if a relative deviation in the surface normal between the presumed corresponding tooth regions is lower than the threshold for surface normal and if a relative deviation in the surface curvature between the presumed corresponding tooth regions is lower than the threshold for surface curvature.

[0062] In an embodiment of the disclosure, a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0063] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0064] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0065] - identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0066] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying those closest tooth regions with a relative deviation in the surface roughness lower than the threshold for the surface roughness,

[0067] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions,

[0068] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model. Aligning of the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions is usually performed on an individual tooth- to-tooth level where regions of a tooth on the first digital three-dimensional model are aligned with corresponding regions of that same tooth on the second digital three-dimensional model.

[0069] Alignment of the digital three-dimensional representations based on surface roughness filtering is advantageous, as it may be possible to exclude tooth regions affected by dental plaque from the alignment process. Tooth regions in the first and / or second digital three-dimensional model representing parts of the dentition with dental plaque are characterized by certain surface roughness and if the relative deviation in the surface roughness between them is higher than the threshold for surface roughness, than those regions can be excluded from the alignment .

[0070] It may be assumed that regions of tooth surfaces without dental plaque, ones that are desirable for alignment purposes, are not rough and that those tooth surfaces have a low average roughness value. In an example, a threshold value for surface roughness may be set to a value up to 2.5 micrometers, more preferably up to 2 micrometers such as 1.6 micrometers.

[0071] In an embodiment of the disclosure, a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0072] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions , receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0073] - identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0074] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying those closest tooth regions with a relative deviation in the surface curvature lower than the threshold for the surface curvature,

[0075] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0076] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0077] Alignment of the digital three-dimensional models based on surface curvature filtering may be advantageous as it may be possible to exclude regions affected by tooth wear from the alignment process. Tooth wear may be characterized by strong change in the surface curvature in regions where dentin is lost in abrasion and / or abfraction. By filtering the presumed corresponding tooth regions of the first and second digital three-dimensional model based on the relative deviation in the surface curvature, and subsequently aligning the three- dimensional models based on thereby obtained actual corresponding tooth regions, it is ensured that tooth regions affected by tooth wear are excluded from the alignment.

[0078] In an embodiment of the disclosure, a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising: - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0079] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0080] - identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0081] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying those closest tooth regions with a relative deviation in the color lower than the threshold for the color,

[0082] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0083] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0084] Performing alignment of the digital three-dimensional models based on the color filtering is advantageous as it may be possible to identify and exclude regions affected by severe tooth wear from the alignment process. For example, a tooth affected by severe tooth wear may change its color from white in enamel region, to more yellow in dentin region. Color may be evaluated as numerical values in RGB color space.

[0085] In an embodiment of the disclosure, a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0086] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth Y1 regions

[0087] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0088] - identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0089] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying those closest tooth regions with a relative deviation in the surface normal lower than the threshold for the surface normal and with a relative deviation in the surface curvature lower than the threshold for the surface curvature,

[0090] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0091] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0092] All presented examples of alignment methods may further be improved if an initial alignment is performed between the first and second digital three-dimensional representation. The aim of the initial alignment is to provide a starting point for the subsequent alignment in which minimizing of distances between the identified corresponding tooth regions will be implemented. By performing the initial alignment, the subsequent alignment step can converge in fewer iterations. Thus, by performing the initial alignment, a higher probability exists that the presumed corresponding tooth regions will be the actual corresponding tooth regions in the context of the disclosure. As a result, lower number of iterations may be required for the algorithm to converge and for the digital three-dimensional models to become aligned .

[0093] The method according to the disclosure may comprise performing the initial alignment of the first digital three-dimensional model and the second digital three-dimensional model.

[0094] The initial alignment may, in one example, be performed by obtaining a best-fit transformation between the first digital three-dimensional model and the second digital three-dimensional model. This may be understood as a jaw-to-jaw alignment which is a rigid transformation in which individual teeth cannot move.

[0095] The best-fit transformation may comprise overlapping centroids of the corresponding teeth of the first digital three- dimensional model and the second digital three-dimensional model .

[0096] Corresponding teeth relate to teeth with the same Universal Numbering System (UNN) notation of the first digital three- dimensional model and the second digital three-dimensional model. Therefore, corresponding tooth regions relate to geometrically matching tooth regions of the corresponding teeth.

[0097] In another example, the initial alignment may comprise aligning tooth poses of the corresponding teeth of the first and second digital three-dimensional model. A tooth pose is a coordinate system originating at the tooth centroid and coinciding with principal tooth axes. Aligning tooth poses may thus impose the starting point for subsequent alignment where minimizing of the distance between the identified corresponding tooth regions will be implemented.

[0098] By comparing the best-fit transformation and the transformation obtained by aligning tooth poses of the corresponding teeth, it may be possible to evaluate a relative tooth motion between the corresponding teeth. This relative tooth motion may be a measure of the progress of an orthodontic treatment.

[0099] In one example, minimizing distances between the identified corresponding tooth regions may be performed by using an Iterative Closest Point (ICP) algorithm. The ICP algorithm is an iterative algorithm and may comprise a number of iterations of identifying corresponding tooth regions and minimizing the distances between them, until the algorithm converges to a desired level of accuracy. However in some examples, a genetic method, an accelerated random sampling method or a random sample consensus (RANSAC) method may be employed as alternatives to the ICP algorithm.

[0100] The method according to the disclosure may further comprise evaluating reliability of the alignment. The reliability of the alignment may be evaluated, for example, by computing a ratio of the area of corresponding tooth regions to a total tooth surface area and comparing the obtained ratio to a threshold. If the ratio is above a certain threshold, it may be considered that the overall alignment is based on a sufficient surface area of teeth of the digital three-dimensional models, and that it is therefore reliable.

[0101] In an embodiment according to the disclosure, a computer- implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0102] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0103] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0104] - optionally performing the initial alignment between the first digital three-dimensional model and the second digital three- dimensional model,

[0105] - obtaining closest tooth regions by identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0106] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying those closest tooth regions with a relative deviation in the surface feature lower than the threshold for the surface feature,

[0107] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0108] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0109] Aligning of the first digital three-dimensional model and the second digital three-dimensional model by minimizing the distance between the identified corresponding tooth regions is usually performed on a tooth-to-tooth level. In the case of alignment of anterior teeth, given they occasionally lack distinctive geometric features to align-by, improvements may be achieved by applying a neighbor anchoring technique which considers not only a tooth-to-be-aligned, but also the two closest neighboring teeth to the tooth-to-be-aligned . In this way, the area used for the alignment may be enlarged and any area affected by geometrical change is likely to be much smaller compared to the enlarged area. Additionally, more distinctive geometric features can serve as the corresponding tooth regions used in the alignment. Geometrical change may occur due to tooth wear for or plaque accumulation example, as mentioned earlier.

[0110] However, the neighbor anchoring technique should be used in cases where no significant relative tooth motion of the anterior teeth is present, otherwise alignment may be inaccurate. The relative tooth motion may be an effect of an orthodontic treatment. The significant relative tooth motion may be detected by comparing aligned tooth poses transformation of the corresponding teeth with the best-fit transformation of the first digital three-dimensional model and the second digital three-dimensional model. If the results of this comparison are greater than predefined thresholds for translation and / or rotation, then significant tooth motion may be present, for example due to an ongoing orthodontic treatment.

[0111] The significant tooth motion caused by the orthodontic treatment may be identified if a translation magnitude, resulting from the comparison between the aligned tooth poses transformation of the corresponding teeth and the best-fit transformation, is greater than one millimeter and / or if a rotation magnitude, resulting from the same comparison, is over 5° expressed as Euler angles.

[0112] In general, the threshold for translation may be in a range of 0.1 millimeter to 1 millimeter while the threshold for rotation may be in a range of 1° to 5°, expressed as Euler angles.

[0113] Thus, in an embodiment of the disclosure a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0114] - receiving the first digital three-dimensional model of the patient's dentition,

[0115] - receiving the second digital three-dimensional model of the patient's dentition - aligning the tooth poses of the corresponding teeth of the first and the second three-dimensional model to obtain tooth pose transformation,

[0116] - computing the best-fit transformation between the first and the second three-dimensional model,

[0117] - comparing the tooth pose transformation to the best-fit transformation to calculate the translation magnitude and the rotation magnitude,

[0118] - if the translation magnitude is lower than the threshold for translation and if the rotation magnitude is lower than the threshold for rotation, then

[0119] - aligning a tooth of the first digital three-dimensional model and the corresponding tooth of the second digital three- dimensional model, wherein aligning is based on the tooth and the two nearest neighboring teeth of the tooth.

[0120] In an embodiment a computer-implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0121] - receiving the first digital three-dimensional model of the patient's dentition,

[0122] - receiving the second digital three-dimensional model of the patient's dentition,

[0123] - performing the initial alignment of the first and second digital three-dimensional models,

[0124] - based on the initial alignment, determining that the movement between corresponding teeth of the first digital three- dimensional model and the second digital three-dimensional model is below the threshold for translation and the threshold for rotation,

[0125] - identifying the corresponding tooth regions of the first digital three-dimensional model and the second digital three- dimensional with a relative deviation in the surface parameter lower than the threshold for the surface parameter, and

[0126] - aligning the tooth of the first digital three-dimensional model and the corresponding tooth of the second digital three- dimensional model based on the corresponding tooth regions, wherein surface to be aligned comprises the tooth and the two nearest neighboring teeth of the tooth.

[0127] Once the first digital three-dimensional model and the second digital three-dimensional model are aligned, differences between the corresponding teeth may be computed. The differences may refer to geometric differences, such as distances between surfaces of the corresponding teeth and / or volumetric differences between the corresponding teeth. These differences may quantify dental conditions such as tooth wear, dental plaque or tartar, dental recession. When the differences between the aligned corresponding teeth are computed, the differences may be visualized on the first digital three-dimensional model, the second digital three-dimensional model and / or on the aligned three-dimensional models.

[0128] In an embodiment according to the disclosure, a computer- implemented method for aligning digital three-dimensional models of the patient's dentition is disclosed, the method comprising:

[0129] - receiving the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0130] - receiving the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0131] - optionally performing the initial alignment between the first digital three-dimensional model and the second digital three- dimensional model

[0132] - obtaining closest tooth regions by identifying, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions,

[0133] - identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, from the closest tooth regions, those closest tooth regions with a relative deviation in the surface features lower than the thresholds for the surface features,

[0134] - aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0135] - displaying the aligned first digital three-dimensional model and the second digital three-dimensional model. The surface features may comprise any two or more of the surface normal, the surface curvature, the surface roughness, the color, the spin images .

[0136] Brief description of the figures

[0137] Aspects of the disclosure may be best understood from the following detailed description taken in conjunction with the accompanying figures. The figures are schematic and simplified for clarity, and they just show details to improve the understanding of the claims, while other details are left out. The individual features of each aspect may each be combined with any or all features of other aspects. These and other aspects, features and / or technical effects will be apparent from and elucidated with a reference to the illustrations described hereinafter in which: Figure 1 illustrates a first digital three-dimensional model and a second three-dimensional model of patient's teeth.

[0138] Figures 2A and 2B illustrate the corresponding teeth of the first and second digital three-dimensional models with a highlighted pair of corresponding regions.

[0139] Figure 3 illustrates a method according to an embodiment of disclosure .

[0140] Figure 4 illustrates a method according to a further embodiment of disclosure .

[0141] Figure 5 illustrates a method according to a yet further embodiment of disclosure.

[0142] Figure 6 illustrates determining relative tooth motion according to an embodiment .

[0143] Figure 7 illustrates a dental scanning system.

[0144] Figure 8 illustrates a neighbor anchoring principle for tooth alignment .

[0145] Figure 9 illustrates an example of a computer architecture capable of carrying out a method according to the disclosure.

[0146] Detailed description

[0147] In the following description, reference is made to the accompanying figures, which show by way of illustration how the invention may be practiced.

[0148] Figure 1 illustrates a first digital three-dimensional model 101 of patient's teeth and a second three-dimensional model 102 of the same patient' s teeth. The first digital three-dimensional model 101 may be obtained by scanning the patient's teeth, with a three-dimensional (3D) scanner, for example with an intraoral 3D scanner 725, at a first time such as a first visit to a dental clinic. The second three-dimensional model 102 may be obtained by scanning the patient's teeth a second time, later than the first time, for example during a subsequent visit to the dental clinic.

[0149] In between the first visit and the second visit to the dental clinic, the patient may have suffered from a dental condition such as tooth wear, illustrated as dark areas on the second three-dimensional model 102. These dark areas represent a loss in tooth material due to tooth wear. Thus, a difference in geometry between the first and second digital three-dimensional models 101, 102 exists. The difference in geometry may be a result of change in anatomy of teeth due to tooth wear. Figure 1 illustrates the patient's upper jaw, however the first and / or second digital three-dimensional models 101, 102 may comprise additionally or alternatively the lower jaw.

[0150] For accurate quantification of the geometric change in the teeth of the patient, it may be desired to align the first digital three-dimensional model 101 and the second three-dimensional model 102. This may be required for diagnostic purposes, such as determining presence or progression of dental conditions such as tooth wear, caries, dental plaque, tooth cracks, gingivitis and / or dental recession. Aligning the two three-dimensional models 101, 102 may be required to track progress of an orthodontic treatment, to plan an orthodontic treatment and / or to plan for other dental procedures. Alignment may be understood as rigid alignment of the two digital three-dimensional models 101, 102 on a tooth-to- tooth level, in the same three- dimensional space. The first and second digital three-dimensional models 101, 102 may have corresponding teeth 103, 103' . Corresponding teeth may relate to teeth with the same Universal Numbering System notation (UNN) on the first and second digital three-dimensional model 101, 102. For example, molar 103 on the first digital three-dimensional model 101 has a corresponding molar 103' on the second digital three-dimensional model 102 with the same UNN notation. Thereby, teeth 103 and 103' illustrated in Figure 1 are examples of corresponding teeth.

[0151] Alignment of the two three-dimensional models 101, 102 may be based on alignment of the corresponding teeth 103, 103' . In order to achieve accurate alignment of the two digital three- dimensional models 101, 102, only areas of the corresponding teeth 103, 103' with matching geometry should be considered. These areas are anatomically corresponding areas, unaffected by different dental conditions and may be referred to as corresponding tooth regions. Change in geometry of teeth of the digital three-dimensional models 101, 102 may be caused for example by tooth wear resulting in loss of tooth material. Another example of the change in geometry is accumulation of bacterial material on a tooth due to dental plaque. As illustrated in Figure 1, the molar 103' on the second three- dimensional model 102 has been affected by tooth wear, and the area 104 represents the loss of dental material.

[0152] Figures 2A and 2B illustrate the corresponding teeth 103, 103' with a pair of corresponding tooth regions 200. The corresponding tooth regions 200 may be understood as those regions of the corresponding teeth 103, 103' with a relative deviation in one or more surface features lower than a threshold for the respective surface feature. This means that the difference in value of the surface feature for the corresponding tooth regions 200 is lower than the threshold for the surface feature, or if more surface features are considered, then the differences in values of the more surface features are all lower than the respective thresholds. Therefore, maximum deviation for each surface feature of corresponding tooth regions 200 is defined by a specific threshold.

[0153] By observing that the relative deviations in the selected number of surface features of the regions 200 of the corresponding teeth 103, 103' are each lower than the respective threshold, it may be concluded that the regions 200 are the corresponding tooth regions 200 of the corresponding teeth 103, 103' . The corresponding tooth regions 200 have matching geometry and may be used in the alignment of the first and second digital three- dimensional models 101, 102. These regions represent unchanged parts of the patient's dentition during the time in-between obtaining the first digital three-dimensional model 101 and the second digital three-dimensional model 102.

[0154] A surface feature of a tooth region may be any one of: a surface normal representing local orientation of the region, a surface curvature of the tooth region, a surface roughness of the tooth region, a color of the tooth region, a point feature histogram or spin images of the tooth region. The tooth region may be a single point of a point cloud representing a tooth, or it may comprise a plurality of points of the tooth point cloud. The tooth region may alternatively be a single facet of a tooth mesh or a plurality of neighboring facets of the tooth mesh.

[0155] If the relative deviation in at least one surface feature, if more surface features are considered, between presumed corresponding tooth regions is above the threshold for that surface feature, then it may be concluded that significant change in geometry between the presumed corresponding tooth regions exists. Those regions may be disregarded for the purpose of aligning the first digital three-dimensional model 101 and the second digital three-dimensional model 102.

[0156] The surface curvature may measure a local change of direction of a tooth region and may be useful for identifying a peak, a pit, a saddle, a ridge and / or a flat part of the tooth region. The surface roughness usually measures a variation in height over the tooth region. The color of the tooth region may be obtained from color information comprised in the first and / or the second digital three-dimensional model 101, 102. Spin images may relate to two-dimensional images of neighboring regions viewed from a local perspective.

[0157] Accordingly, the threshold for a selected surface feature may relate to any one of: a threshold for the surface normal, a threshold for the surface curvature, a threshold for the surface roughness, a threshold for the color, a threshold for point feature histogram, a threshold for the spin images. In general, the threshold may be a tunable parameter derived from previously aligned digital representations.

[0158] Figure 2A shows the corresponding teeth 103, 103' overlaid with a heatmap representing mean (H) surface curvature. Normalized values of the surface curvature may be indicated on a bar 201. Instead of the mean (H) surface curvature, Gaussian (K) surface curvature may be observed. In one example, the threshold for the surface curvature may be set to 0.3 1 / mm for the mean (H) surface curvature. Points 200 on Figure 2A form one pair of corresponding tooth regions, one tooth region being on the first digital three-dimensional model 101 and a second tooth region being on the second digital three-dimensional model 102. The points 200 are characterized by the relative deviation in the surface curvature below the threshold for the surface curvature.

[0159] Figure 2B shows the corresponding teeth 103, 103' with the corresponding tooth regions 200. On the tooth 103, the surface normal n is illustrated for the tooth region 200. On the tooth 103' , a corresponding surface normal n' is illustrated. The angle 0 between the surface normal n and the surface normal n' is a measure of deviation in the surface normal of these corresponding tooth regions 200. The threshold for the surface normal relates to the angle 0 between two surface normals and may be set to 30°. Thus, the corresponding tooth regions would be those regions with surface normals deviating from each other less than 30°.

[0160] Figures 2A and 2B show only one pair of the corresponding tooth regions 200, however multiple pairs of such regions may be identified prior to aligning the corresponding teeth 103, 103' .

[0161] In an embodiment, more than one surface feature may be evaluated to determine the corresponding tooth regions 200, such as both the surface curvature and the surface normal. Thereby, the corresponding tooth regions 200 may be characterized by a relative deviation in the surface curvature less than the threshold for surface curvature, and a relative deviation in the surface normal less than the threshold for surface normal.

[0162] Figure 3 illustrates a method 300 for aligning the first digital three-dimensional model 101 and the second digital three- dimensional model 102, according to an embodiment. Steps 301 and 302 illustrate receiving the first and second digital three- dimensional models 101, 102 by a processor. The two 3D models 101, 102 may be received simultaneously or sequentially. The two 3D models 101, 102 may be received, by the processor, in a stereolithography (STL) format or any other format that can be converted into a digital 3D representation.

[0163] In step 303, an initial alignment of the first digital three- dimensional model 101 and the second digital three-dimensional model 102 may optionally be performed. The initial alignment may be beneficial to identify a starting point from which a subsequent alignment algorithm can converge faster compared to the scenario where no initial alignment is performed.

[0164] The initial alignment may be performed in several manners . In an example, a best-fit transformation may be performed between the two digital 3D models 101, 102 in which centroids of all of the corresponding teeth are overlapped. The best-fit transformation is a rigid transformation which, when applied to teeth centroids of the first digital three-dimensional model 101, minimizes the sum of squared distances to teeth centroids of the second digital three-dimensional model 102. This best-fit transformation may be regarded as a jaw-to-jaw alignment as it is computed on the jaw level and not on the level of individual teeth. The obtained jaw-to-jaw alignment may be fine-tuned by performing an Iterative Closest Point (ICP) method considering selected teeth, for example molars, of the two 3D models 101, 102.

[0165] In another example, the initial alignment may be performed by aligning tooth poses of all of the corresponding teeth 103, 103' of the first and second digital three-dimensional model 101, 102. A tooth pose is a coordinate system originating at the tooth centroid with axes corresponding to the principal axes of the tooth. Aligning tooth poses may thus impose the starting point for subsequent alignment step. In step 304, the corresponding tooth regions 200 of the first digital three-dimensional model 101 and the second digital three-dimensional model 102 are identified. This may be performed by comparing a first plurality of tooth regions comprised in the first digital three-dimensional model 101 with a second plurality of tooth regions comprised in the second digital three-dimensional model 102 and identifying, for each tooth region of the first plurality of tooth regions, a tooth region of the second plurality of tooth regions with a minimal relative deviation in the surface feature, for example surface normal or surface curvature. Such obtained pairs of tooth regions may be referred to as the corresponding tooth regions 200.

[0166] Thus, in the method according to an embodiment, the selected one or more surface features may be used to identify the corresponding tooth regions 200 representing those parts of the digital three-dimensional models 101, 102 with unchanged geometry. The corresponding teeth 103, 103' may comprise multiple pairs of individual corresponding tooth regions.

[0167] In step 305 aligning of the first digital three-dimensional model 101 and the second digital three-dimensional model 102 by minimizing distances between the identified corresponding tooth regions 200 is performed. This alignment step may be understood as tooth- to- tooth alignment because it may be performed on individual tooth level. Alignment of the identified corresponding tooth regions 200 may be achieved by running the Iterative Closest Point (ICP) algorithm. The ICP algorithm is an iterative algorithm and may comprise a number of iterations of identifying corresponding tooth regions 200 and minimizing the distances between the identified tooth regions 200, until the algorithm converges to a desired result. In further examples, using a genetic method, an accelerated random sampling method or a random sample consensus (RANSAC) method may be used instead of the ICP algorithm.

[0168] In step 306 a reliability of the alignment may be optionally performed. The reliability may be evaluated, for example, by evaluating a ratio of the total area of corresponding tooth regions 200 to a total tooth surface area of the aligned teeth 103, 103' . If the reliability of alignment is not satisfactory then the alignment may continue with further iterations. A threshold for the ratio may be set up to a value of 50%.

[0169] More generally, the threshold for the ratio may be set up to a value in range of 40%-80%. If the alignment does not converge, then a message may be presented to the user, for example as a visual message, audio message or audio-visual message informing the user of the alignment reliability.

[0170] In step 307, the aligned first and second digital three- dimensional models 101, 102 are displayed in a common three- dimensional space.

[0171] Figure 4 illustrates a method 400 for aligning the first digital three-dimensional model 101 and the second digital three- dimensional model 102, according to a further embodiment.

[0172] Steps 401 and 402 are equivalent to steps 301 and 302 described with respect to Figure 3. Moreover, optional step 403 of performing the initial alignment of the first and second digital three-dimensional model 101, 102 is equivalent to step 303 of Figure 3.

[0173] In step 408 of Figure 4, presumed corresponding tooth regions of the first digital three-dimensional model 101 and the second digital three-dimensional model 102 are identified. The presumed corresponding tooth regions may be found by identifying, for each tooth region of the first plurality of tooth regions comprised in the first digital three-dimensional model 101, a closest tooth region of the second plurality of tooth regions comprised in the second digital three-dimensional model 102. Thus, the presumed corresponding tooth regions are those tooth regions with shortest distances between the first and second digital three-dimensional models 101, 102. These presumed corresponding tooth regions may or may not be the actual corresponding tooth regions 200 to be used in the subsequent alignment step. A further step may therefore be required, to filter out the corresponding tooth regions 200 from the identified plurality of presumed corresponding tooth regions.

[0174] The presumed corresponding tooth regions may also be referred to as the closest tooth regions.

[0175] In step 404 filtering may be applied to the previously identified presumed corresponding tooth regions. The filtering may be based on the one or more surface feature characterizing the presumed corresponding tooth regions.

[0176] The presumed corresponding tooth regions may be filtered with respect to a relative deviation in the surface normal, surface curvature, surface roughness, color, point feature histograms and / or spin images. In one example a filter may be set up based on the surface normal and the surface curvature. The threshold for surface normal may be set to 30° and a threshold for surface curvature may be set to 0.3 1 / mm for the mean surface curvature. Thereby, those presumed corresponding tooth regions with relative deviation in both the surface curvature and the surface normal lower than above-mentioned thresholds will be selected as the corresponding tooth regions 200.

[0177] The mentioned threshold for surface normal of 30° and threshold for surface curvature of 0.3 1 / mm are exemplary, and these values may be best understood as upper values of the respective thresholds. Thus, the threshold for surface normal may be set to a value in range of 5° to 30°. The threshold for surface curvature may be set to a value up to 0.3 1 / mm for the mean surface curvature.

[0178] In step 405 the first digital three-dimensional model 101 and the second digital three-dimensional model 102 are aligned based on the previously identified corresponding tooth regions 200. This step may be analogous to step 305 of Figure 3.

[0179] Additionally, steps 406 and 407 are analogous to steps 306 and 307 of Figure 3.

[0180] Figure 5 illustrates a method 500 for aligning the first digital three-dimensional model 101 and the second digital three- dimensional model 102, according to a further embodiment.

[0181] Steps 501 and 502 are equivalent to steps 301 and 302 described with respect to Figure 3.

[0182] In step 503, the initial alignment of the first digital three- dimensional model 101 and the second digital three-dimensional model 102 is performed. The initial alignment may be useful to quantify relative tooth motion between the corresponding teeth 103, 103' . How the relative tooth motion may be quantified is further described with respect to Figure 6.

[0183] In step 509 it is estimated whether the relative tooth motion between any of the corresponding teeth is below a predetermined threshold for tooth motion. If the relative tooth motion is below the threshold for tooth motion, then the subsequent alignment of the individual corresponding teeth may be improved by considering a left and a right neighboring tooth to the individual tooth to be aligned. In this way, tooth surface used for alignment of individual teeth is enlarged. This is particularly beneficial when aligning anterior teeth due to lack of distinctive geometric features of the anterior teeth to align on .

[0184] In step 504, the corresponding tooth regions 200 are identified using one or more surface features. This identification may be according to the description of step 304 of Figure 3, or according to steps 408 and 404 of Figure 4.

[0185] In step 505, aligning of the first digital three-dimensional model 101 and the second digital three-dimensional model 102 is performed. When computing the tooth- to- tooth alignment in this case of method 500, the nearest two neighboring teeth of the tooth- to-be-aligned may be considered. Thus, the surface to-be- aligned may be increased, resulting in a more accurate tooth-to- tooth alignment. As previously mentioned, this may be of particular relevance if the tooth-to-be-aligned is an anterior tooth. Alignment of the step 505 may utilize any of the alignment algorithms previously disclosed, such as the ICP algorithm, the genetic method, the accelerated random sampling method or the random sample consensus (RANSAC) method.

[0186] Optional step 506 of evaluating reliability of the alignment and step 507 of displaying the aligned digital three-dimensional models 101, 102 are analogous to steps 306 and 307 of Figure 3.

[0187] Figure 6 illustrates how the relative tooth motion between corresponding teeth such as teeth 103 and 103' can be estimated. This estimate may be based on comparison of the best-fit transformation 604 between the first digital three-dimensional model 101 and the second digital three-dimensional model 102, and the transformation 605 obtained by aligning tooth poses of all of the corresponding teeth of the two digital three- dimensional models 101, 102.

[0188] The best-fit transformation 604 between the first digital three- dimensional model 101 and the second digital three-dimensional model 102, and the transformation 605 obtained by aligning tooth poses of all of the corresponding teeth of the two digital three-dimensional models 101, 102 may be used for the initial alignment of the two digital three-dimensional models. A characteristic of the best-fit transformation 604 is that individual teeth cannot move and instead the alignment is performed on a jaw-level and referred to as a jaw-to- aw alignment. On the contrary, a characteristic of the transformation 605 obtained by aligning the tooth poses is that individual teeth are allowed to move.

[0189] These two transformations 604, 605 may be compared, as illustrated in block 601. In the comparison, a rigid misfit in terms of x, y, z displacement is calculated between the two transformations, resulting in a translation value and a rotation value. The rotation value represents angular displacement about the x, y, z axes between the two transformations. This comparison may be computed for each tooth.

[0190] The computed translation and rotation need to be compared to their respective thresholds for translation and rotation, as indicated in blocks 602 and 603. To establish that the relative tooth motion is below the predefined threshold, computed translation should be below the threshold for translation and the computed rotation should be below the threshold for rotation .

[0191] Figure 7 illustrates a dental scanning system 700 which may comprise a computer 710 capable of carrying out any method of the disclosure. The computer may comprise a wired or a wireless interface to a server 715, a cloud server 720 and the intraoral scanner 725. The intraoral scanner 725 may be capable of recording the scan data comprising geometrical information, natural color information and / or fluorescence information associated with the patient' s dentition. The intraoral scanner 725 may be equipped with various modules such as a fluorescence module or an infrared module and thereby capable of capturing information relevant for diagnosing dental conditions such as caries, tooth cracks, gingivitis and / or plaque.

[0192] The dental scanning system 700 may comprise a data processing device configured to carry out the method according to one or more embodiments of the disclosure. The data processing device may be a part of the computer 710, the server 715 or the cloud server 720.

[0193] The dental scanning system 700 may comprise the data processing device configured to:

[0194] - receive the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0195] - receive the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0196] - identify, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions to obtain closest tooth regions, - identify corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, from the closest tooth regions, those closest tooth regions with the relative deviation in the surface feature lower than the threshold for the surface feature,

[0197] - align the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0198] - display the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0199] A non- transitory computer-readable storage medium may be comprised in the dental scanning system 700. The non- transitory computer-readable medium can carry instructions which, when executed by a computer, cause the computer to carry out the method according to one or more embodiments of the disclosure.

[0200] The non- transitory computer-readable medium can carry instructions which, when executed by a computer, cause the computer to:

[0201] - receive the first digital three-dimensional model of the patient's dentition comprising the first plurality of tooth regions ,

[0202] - receive the second digital three-dimensional model of the patient's dentition comprising the second plurality of tooth regions ,

[0203] - identify, for each tooth region of the first plurality of tooth regions, the closest tooth region of the second plurality of tooth regions to obtain closest tooth regions,

[0204] - identify corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, from the closest tooth regions, those closest tooth regions with the relative deviation in the surface feature lower than the threshold for the surface feature,

[0205] - align the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions, and

[0206] - display the aligned first digital three-dimensional model and the second digital three-dimensional model.

[0207] A computer program product may be embodied in the non-transitory computer-readable storage medium. The computer program product may comprise instructions which, when executed by a computer, cause the computer to perform the method according to any of the embodiments presented herein.

[0208] Figure 8 illustrates a neighbor anchoring principle for tooth alignment. Figure 8 shows the tooth 103 and its corresponding tooth 103' which are to be mutually aligned as indicated by the dashed arrow 801.

[0209] In the case where the corresponding teeth 103, 103' are anterior teeth, such as shown in figure 8, improvement in alignment may be achieved by applying the neighbor anchoring technique which considers not only a tooth-to-be-aligned 103, 103' , but also the two closest neighboring teeth to the tooth- to-be-aligned 103, 103' . The neighboring teeth are shown as shaded teeth on the left and right side of the tooth- to-be-aligned 103, 103' . Due to the anatomical characterists of the anterior teeth they occasionally lack distinctive geometric features to align-by. By considering the neighboring teeth additionally, the area used for the alignment is enlarged and any area affected by geometrical change is likely to be much smaller compared to the enlarged area. In this way, the accuracy of the alignment of teeth 103, 103' is improved. Figure 9 illustrates an example of a computer architecture of a computer capable of carrying out a method according to the disclosure .

[0210] Various components of the computer 710 may communicate via a bus 910.

[0211] The computer 710 may comprise the data processing device 920 (referred to also as a processor or a processing device) . The data processing device 920 may be any central processing unit (CPU) , microprocessor, microcontroller, computational or programmable device or circuit configured for executing instructions to carry out the method of any one or more of the presented embodiments.

[0212] The computer program product 940, comprising the instructions to carry out the method of any one or more of the presented embodiments, may be stored on the data processing device 920. Alternativelly or additionaly, the computer program product 940, comprising the instructions to carry out the method of any one or more of the presented embodiments, may be stored on a computer-readable medium 930, more specifically on a non- transitory computer-readable medium 930. Examples of the computer-readable medium 930 include magnetic storage media such as a magnetic disk or magnetic tape, optical storage media such as an optical disc, optical tape, machine readable bar code, solid state electronic storage devices such as random access memory (RAM) , read only memory (ROM) , or any other physical device or medium configured to store the computer program product 940.

[0213] The computer 710 may further comprise an input / output device 950 such as a keyboard, a touchscreen, a microphone, a mouse, a display, a graphical user interface (GUI) , a loudspeaker etc. The computer 710 may be connected to the server 715, the cloud 720 and / or the intraoral scanner 725 via an interface device 960 which may be a wired and / or wireless communication interface device including Wi-Fi, Bluetooth, LAN, etc.

[0214] It is to be understood that embodiments may be made, other than those mentioned, and structural and functional modifications may be made without departing from the scope of the present invention .

[0215] It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" or "an aspect" or features included as "may" means that a particular feature, structure, or characteristic described in connection with the embodiments is included in at least one embodiment of the disclosure. Furthermore, the particular features, structures or characteristics may be combined as suitable in one or more embodiments of the disclosure. The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects.

Claims

Claims1. A computer-implemented method for aligning digital three- dimensional models of a patient' s dentition, the method comprising :- receiving a first digital three-dimensional model of the patient's dentition comprising a first plurality of tooth regions ;- receiving a second digital three-dimensional model of the patient's dentition comprising a second plurality of tooth regions ;- obtaining closest tooth regions by identifying, for each tooth region of the first plurality of tooth regions, a closest tooth region of the second plurality of tooth regions;- identifying corresponding tooth regions of the first digital three-dimensional model and the second digital three-dimensional model by identifying, from the closest tooth regions, those closest tooth regions with a relative deviation in a surface feature lower than a threshold for the surface feature, wherein the surface feature is a surface normal and wherein the threshold is a threshold for surface normal selected in a range of 5 degrees to 30 degrees;- aligning the first digital three-dimensional model and the second digital three-dimensional model by minimizing distances between the identified corresponding tooth regions; and- displaying the aligned first digital three-dimensional model and the second digital three-dimensional model.

2. The method according to the previous claim 1, further comprising initially aligning the first digital three- dimensional model and the second digital three-dimensional model .

3. The method according to the previous claim 2, wherein initially aligning the first digital three-dimensional model and the second digital three-dimensional model comprises aligning tooth poses of corresponding teeth.

4. The method according to the previous claim 3, wherein initially aligning the first digital three-dimensional model and the second digital three-dimensional model further comprises overlapping the centroids of corresponding teeth.

5. The method according to the previous claim 4, further comprising quantifying a relative tooth motion between corresponding teeth of the first digital three-dimensional model and the second digital three-dimensional model based on the initial alignment.

6. The method according to the previous claim 5, wherein quantifying the relative tooth motion comprises comparing a transformation obtained by aligning tooth poses of corresponding teeth with a transformation obtained by overlapping the centroids of corresponding teeth.

7. The method according to the previous claim 5 or 6, further comprising aligning the corresponding teeth of the first digital three-dimensional model and the second digital three-dimensional model if the relative tooth motion is below a threshold for translation and / or a threshold for rotation, wherein aligning of the corresponding teeth comprises alignment of a tooth-to-be- aligned based on two closest neighboring teeth.

8. The method according to the previous claim 7, wherein the threshold for translation is in a range of 0.1 millimeter to 1 millimeter .

9. The method according to the previous claim 7 or 8, wherein the threshold for rotation is in a range of 1° to 5°, expressed as Euler angles.

10. The method according to any previous claim, further comprising computing a ratio of an area of the corresponding tooth regions to a total surface area of corresponding teeth, and comparing the obtained ratio to a threshold for the ratio.

11. The method according to the previous claim 10, wherein the threshold for the ratio is in range of 40%-80%.

12. The method according to any previous claim, wherein the corresponding tooth regions are identified using a genetic method or accelerated random sampling method or a random sample consensus method.

13. A computer program product comprising instructions which, when executed by a computer, cause the computer to perform the method according to any previous claim 1-12.

14. A non- transitory computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to carry out the method according to any previous claim 1-12.

15. A dental scanning system comprising:- a memory;- a data processing device coupled to the memory, the data processing device configured to carry out the method according to any previous claim 1-12.

Images