DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. EP22185244.5, filed on 2022.07.15.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 13 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
Claim 13 is directed to A computer program product for use in a device which does not fall within at least one of the four categories of patent eligible subject matter recited in 35 U.S.C. 101 (process, machine, manufacture, or composition of matter).
A "computer program product " is not explicitly or deliberately defined to include only the non-transitory embodiments listed on Page 18: “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”. The broadest reasonable interpretation of a claim drawn to a computer readable medium typically covers forms of non-transitory tangible media and transitory propagating signals per se in view of the ordinary and customary meaning of computer readable media. See Subject Matter Eligibility of Computer Readable Media, 1351 OG 212 (26 Jan 2010). See MPEP 2111.01. Signals are nothing but the physical characteristics of a form of energy, and as such is nonstatutory natural phenomena. See, e.g., In re Nuitjen, Docket no. 2006-1371 (Fed. Cir. Sept.20, 2007)(slip. op. at 18)("A transitory, propagating signal like Nuitjen's is not a process, machine, manufacture, or composition of matter.' ... Thus, such a signal cannot be patentable subject matter."). Thus, claims 8 is rejected under 35 U.S.C. 101 because, giving the claims their broadest reasonable interpretation, the claimed "A computer program product for use in a device" is considered to be directed to transitory propagating signals per se and therefore encompasses non-statutory subject matter.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. (FP 7.30.05)This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier.
Such claim limitation(s) is/are: A data processing apparatus comprising means for carrying out the method of Claim 1. In Claim 13.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. (FP 7.30.06)
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-4, 12-15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Feng et al. (US 20210217233 A1, hereinafter Feng) in view of Azernikov et al. (US 20220215531 A1, hereinafter Azernikov).
Regarding Claim 1, Fent teaches a computer-implemented method for tooth pose estimation, the method comprising (Feng, Paragraph [0027], "the Inventors of the present application developed a computer-implemented method for setting a local coordinate system <read on tooth pose> of a tooth 3D digital model based on deep learning. A trained deep learning artificial neural network is used to set the local coordinate system of the tooth 3D digital model."): obtaining a virtual 3D representation representing a patient's dentition (Feng, Paragraph [0034], "a 3D digital model representing a patient's dentition may be obtained by scanning the patient's jaw directly."); segmenting the virtual 3D representation to obtain at least a first segmented tooth representation and a second segmented tooth representation (Feng, Paragraph [0034], "3D digital models representing various teeth respectively may be obtained by segmenting a 3D digital model representing the patient's dentition."); [[ wherein the at least first and second segmented tooth representation represent neighboring teeth in the patient's dentition ]] [[ determining an initial tooth pose for the first segmented tooth representation using a geometric parameter of the first segmented tooth representation and a geometric parameter of the second segmented tooth representation, wherein the geometric parameter of the first segmented tooth representation is a centroid of the first segmented tooth representation and the geometric parameter of the second segmented tooth representation is a centroid of the second segmented tooth representation ]]; wherein each tooth pose is comprised of an origin in which a first axis a second axis and a third axis intersect (Feng, Paragraph [0117], "the origin of the local coordinate system may be the center of the mesh boundary."; [0084], "the coordinate axes of the local coordinate system are represented with x, y and z."); wherein determining the initial tooth pose for the first segmented tooth representation using the geometric parameter of the first segmented tooth representation and the geometric parameter of the second segmented tooth representation comprises determining the first, the second and the third axis of the initial tooth pose for the first segmented tooth representation (Feng, Paragraph [0115], "the y-axis and z-axis of the local coordinate system are determined based on the determined x-axis of the local coordinate system and the first predicted vector.") [[ wherein the first axis the initial tooth pose for the first segmented tooth representation is calculated as a difference between the centroid of the first segmented tooth representation and the centroid of the second segmented tooth representation ]]; obtaining a normalized tooth representation of the first segmented tooth representation by transforming the first segmented tooth representation using the initial tooth pose for the first segmented tooth representation (Feng, Paragraph [0048], "a mean value of coordinates of the vertices of the first digital data set may be calculated, to obtain coordinates of a center point, and then the coordinates of the center point is subtracted from the coordinates of the vertices of the first digital data set to obtain the second digital data set.") and (Feng, Paragraph [0049], "the second digital data set is normalized to obtain a third digital data set."); inputting the normalized tooth representation of the first segmented tooth representation into a trained neural network (Feng, Paragraph [0054], "a first predicted vector corresponding to the z-axis of the local coordinate system is obtained using the trained deep learning artificial neural network based on the third digital data set."); producing an output from the trained neural network, wherein the output comprises a correct tooth pose for the first segmented tooth representation (Feng, Paragraph [0067], "When the label item is the z-axis of the local coordinate system, the first predicted vector is the z-axis of the local coordinate system predicted by the MLP."; [0115], "the y-axis and z-axis of the local coordinate system are determined based on the determined x-axis of the local coordinate system and the first predicted vector.").
But Feng does not explicitly disclose the neighboring-teeth limitation, the centroid-based geometric parameters, or the first-axis-as-centroid-difference computation.
However, Azernikov teaches segmenting the virtual 3D representation to obtain at least a first segmented tooth representation and a second segmented tooth representation, wherein the at least first and second segmented tooth representation represent neighboring teeth in the patient's dentition (Azernikov, Paragraph [0003], "segmenting the labeled 3D digital model to provide a segmented 3D digital model."; [0101], "the neighboring digital tooth bounding regions are adjacent <read on neighboring> to one another. This can be because the digital tooth bounding regions are of neighboring or adjacent teeth"; it is noted segments the dentition into per-tooth regions that are explicitly identified as "neighboring or adjacent teeth); determining an initial tooth pose for the first segmented tooth representation using a geometric parameter of the first segmented tooth representation and a geometric parameter of the second segmented tooth representation (Azernikov, Paragraph [0100], "the computer-implemented method can determine a center of each digital tooth bounding region. The digital tooth bounding region center can be a geometric center <read on centroid> of the digital tooth bounding region"; [0102], "a first bounding box center 808 and a second bounding box center 810 connected together by a first spline 812."; [0165], "determine a center point of one or more labeled regions, construct a best-fit parabola to connect the center points of the one or more labeled regions, construct a plane at each center point such that the plane contains the center point and is orthogonal to the best-fit parabola at the center point"; it is noted by using two distinct per-tooth geometric parameters (the geometric center / centroid of the first segmented tooth region and the geometric center / centroid of the second/neighboring segmented tooth region) to establish a per-tooth spatial reference: a center point that anchors an origin for the first tooth and a directional link to the neighboring tooth's center that fixes an orientation (via the tangent/spline direction between the two centers, and via the orthogonal plane built at the first tooth's center). Anchoring an origin at the first tooth's center and orienting axes using the inter-center direction is, functionally, the determination of an initial tooth pose for the first segmented tooth representation using a geometric parameter of the first segmented tooth representation and a geometric parameter of the second segmented tooth representation); the geometric parameter of the first segmented tooth representation is a centroid of the first segmented tooth representation and the geometric parameter of the second segmented tooth representation is a centroid of the second segmented tooth representation (Azernikov, Paragraph [0100], "The digital tooth bounding region center can be a geometric center <read on centroid> of the digital tooth bounding region in some embodiments"; [0165], "determine a center point of one or more labeled regions, construct a best-fit parabola to connect the center points of the one or more labeled regions"); the first axis the initial tooth pose for the first segmented tooth representation is calculated as a difference between the centroid of the first segmented tooth representation and the centroid of the second segmented tooth representation (Azernikov, Paragraph [0102], "a first bounding box center 808 and a second bounding box center 810 connected together by a first spline 812."; [0101], "splines can be determined by determining tangents in every center point and then constructing the spline in between every two digital tooth bounding region centers"; it is noted draws a directional spline directly between two neighboring tooth centers; the vector from one center to the next-tooth center is, by definition, the difference between the two centroids, which serves as the local mesial–distal directional axis for the first tooth).
Azernikov and Feng are analogous since both are computer-implemented dental processing systems that operate on 3D digital models of a patient's dentition, use trained neural networks for per-tooth tasks, and rely on segmented tooth representations as inputs to downstream geometric reasoning. Feng provided a way of training an artificial neural network to predict axis vectors of a per-tooth local coordinate system (tooth pose) from a centralized + normalized 3D digital tooth model, then deriving the remaining axes via cross products from the predicted vector and an axis derived from the tooth's own geometry. Azernikov provided a way of segmenting the dentition into labeled per-tooth regions using a trained neural network, computing a geometric center (centroid) for each tooth region, and connecting the centers of neighboring teeth with a directional spline so that an inter-tooth direction is available. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Azernikov's per-tooth segmentation, per-tooth centroid (geometric center) computation, and inter-centroid directional link between neighboring teeth into the modified invention of Feng such that, instead of (or in addition to) deriving the first axis via NPCA from a single tooth's surface, the vector from the centroid of the first segmented tooth to the centroid of the neighboring second segmented tooth supplies the first axis of the initial tooth pose, with that initial pose then used to centralize/normalize the first tooth before feeding it to Feng's trained neural network.
Regarding Claim 2, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches wherein the second axis of the initial tooth pose for the first segmented tooth representation is orthogonal to an occlusal surface of the first segmented tooth representation (Feng, Paragraph [0033], "a vector perpendicular <read on orthogonal> to an occlusal surface may be taken as the Z-axis of the world coordinate system <read on second axis of the initial tooth pose>, a line passes through cusps of two No. 6 teeth may be taken as the direction of X-axis, and the Y-axis may be determined based on these two coordinate axes."; [0029], "the computer-implemented method 100 for setting a local coordinate system of a tooth 3D digital model is based on Multi-Layer Perceptron"; [0117], "the origin of the local coordinate system may be the center of the mesh boundary."; it is noted the occlusal-perpendicular axis reads on the claim's second axis of the initial tooth pose that is orthogonal to the occlusal surface).
Azernikov also teaches wherein the second axis of the initial tooth pose for the first segmented tooth representation is orthogonal to an occlusal surface of the first segmented tooth representation (Azernikov, Paragraph [0091], "The occlusal direction is a normal <read on orthogonal> to an occlusal plane and the occlusal plane can be determined for the digital model using any technique known in the art."; it is noted deriving the occlusal direction as the normal of the occlusal plane, confirming that an axis orthogonal to the occlusal surface is a routinely available per-tooth reference direction).
As explained in rejection of claim 1, the obviousness for combining of geometry representation of Azernikov into Feng is provided above.
Regarding Claim 3, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches wherein the third axis the initial tooth pose for the first segmented tooth representation is obtained as a cross product between the first axis of the initial tooth pose for the first segmented tooth representation and the second axis of the initial tooth pose for the first segmented tooth representation (Feng, Paragraph [0116], "the cross product between the determined x-axis and y-axis of the local coordinate system may be calculated and taken as the z-axis <read on third axis> of the modified local coordinate system; [0141], "the cross product of the first coordinate axis and second coordinate axis is taken as the third coordinate axis of the local coordinate system.");
Regarding Claim 4, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches further comprising adapting the first segmented tooth representation by reducing a number of vertices representing the first segmented tooth representation (Feng, Paragraph [0037], "a 3D digital model may be simplified before being input to an artificial neural network, such that the number of vertices/facets thereof is equal to a predetermined value N. In one embodiment, the predetermined value may be set to 2048 or 1024, namely, the number of vertices of the simplified 3D digital model equals to 2048 or 1024.";[0038], "the first 3D digital model may be simplified using an edge contraction algorithm with Quadric Error Metrics as a cost."; [0044], "a merge cost may be calculated for each pair of adjacent vertices, then a set of vertex pairs with a minimum merge cost may be iteratively selected for contraction, and errors of all relevant edges may be updated. The simplified vertex set may be obtained based on the calculation of the Q matrix.");
Regarding Claim 12, it recites limitations similar in scope to the limitations of claim 1, but in an apparatus. As shown in the rejection, the combination of Feng and Azernikov disclose the limitations of claims 1. Additionally, the combination of Feng discloses an apparatus that maps to Paragraph [0027], [0004], (Feng, Paragraph [0027], "computer-implemented method"; [0010], “the computer-implemented method for setting a local coordinate system of a tooth 3D digital model”) and by Azernikov (Azernikov, Paragraph [0045], “For purposes of this description, certain aspects, advantages, and novel features of the embodiments of this disclosure are described herein. The disclosed methods, apparatus, and systems should not be construed as being limiting in any way”). Thus, Claim 13 is met by the combination of Feng and Azernikov according to the mapping presented in the rejection of claims 1, given the method corresponds to the system.
Regarding Claim 13, it recites limitations similar in scope to the limitations of claim 1 and the combination of Feng and Azernikov teaches all the limitations as of Claim 1. And Azernikov discloses these features can be implemented on a computer program (Azernikov, Paragraph [0005], “A non-transitory computer readable medium storing executable computer program instructions to segment a digital model, the computer program instructions can include receiving a 3D digital model”).
Regarding Claim 14, it recites limitations similar in scope to the limitations of claim 1 and the combination of Feng and Azernikov teaches all the limitations as of Claim 1. And Azernikov discloses these features can be implemented on a non-transitory computer-readable medium (Azernikov, Paragraph [0005], “A non-transitory computer readable medium storing executable computer program instructions to segment a digital model, the computer program instructions can include receiving a 3D digital model”).
Regarding Claim 15, the combination of Feng and Azernikov teaches the method in Claim 1, plus an intraoral scanner from Azernikov further teaches an intraoral scanner (Azernikov, Paragraph [0048], "The digital model 100 can also be generated by intraoral scanning of the patient's dentition"), and the computer/data-processing-device infrastructure (Azernikov, Paragraph [0004], "A system to digitally segment teeth in a digital model includes a processor, a computer-readable storage medium including instructions executable by the processor"). The recited "server" and "cloud server" are conventional networked-system components understood by one of ordinary skill in the art to be readily combinable with such a system. Thus, Claim 15 is met by the combination of Feng and Azernikov according to the mapping presented in the rejection of claims 1, given the method corresponds to the system.
Claim(s) 5, 7, 8, 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Feng et al. (US 20210217233 A1, hereinafter Feng) in view of Azernikov et al. (US 20220215531 A1, hereinafter Azernikov) as applied to Claim 1 above and further in view of Claessen et al. (US 20210174543 A1, hereinafter Claessen).
Regarding Claim 5, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination does not explicitly disclose but Claessen teaches wherein the trained neural network is a PointNet neural network (Claessen, Paragraph [0077], "a 3D deep neural network may be used that is capable of determining the canonical pose <read on correct tooth pose> of optical scan data (a 3D point cloud) directly based on the point cloud data. An example of such network are multi-layer perceptron (MLP) based deep neural network. MPL deep neural network architectures include PointNet (Qi, C. R., et al. : Pointnet: Deep learning on point sets for 3d classication and segmentation. Proc. Computer Vision and Pattern Recognition (CVPR), IEEE 1(2), 4 (2017)) or PointCNN"; "These MLP deep neural networks are capable of directly processing points of a point cloud. Such neural networks may be trained to determine the canonical pose of a 3D dental object directly from a point cloud representation of the 3D dental object.").
Claessen and Feng are analogous since both train MLP-based deep neural networks on per-tooth 3D representations of a patient's dental geometry to predict per-tooth pose-related outputs (axis vectors of the local coordinate system in Feng, canonical pose of the 3D dental object in Claessen); both networks share the same MLP foundation that PointNet is built upon. Feng provided a way of using a per-tooth MLP-based neural network to predict an axis vector of the per-tooth local coordinate system (tooth pose), and expressly invited substitution with other suitable deep-learning networks; Claessen provided a way of implementing that MLP-based pose-determination network as a PointNet that directly processes the points of a point-cloud representation of the 3D dental object. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Claessen's PointNet architecture into the modified invention of Feng (in view of Azernikov) such that the trained neural network that consumes the centroid-centered, normalized per-tooth representation and outputs the correct tooth pose is implemented as a PointNet neural network — a simple substitution of one known MLP-family architecture for another in the same predictive role Feng already describes. The motivation is to supply the specific technical reason for choosing PointNet for this task: PointNet is an MLP-based network that can directly process the points of a point cloud and can be trained to determine the canonical (correct) pose of a 3D dental object from such input.
Regarding Claim 7, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination does not explicitly disclose but Claessen teaches wherein the output of the trained neural network comprises a 3-dimensional translation vector comprising a translation amount for the first segmented tooth representation (Claessen, Paragraph [0158], “During the training effort, the internal parameters of 3D deep neural network 1114 may be optimized towards the network providing results of a sufficiently high accuracy”; [0100], "The position of the origin of the canonical coordinate system (in terms of a translation vector in the space of the first coordinate system) may be obtained by determining a prediction of the canonical coordinates of the centre of a voxel representation that is offered to the input of the 3D deep learning network. These coordinates may be determined based on e.g. the average or median value of predicted x′ values of the first 3D voxel map, y′ values of the second 3D voxel map and z′ values of the third 3D voxel map. A translation vector may be determined based on the predicted canonical coordinates"; [0156], "System 1100 may be employed to directly predict transformation parameters, e.g. applicable 3D rotations, 3D translations and 3D scaling").
Claessen and Feng are analogous since both train deep neural networks on per-tooth 3D dental representations to produce pose-related outputs (axis vectors in Feng, canonical-pose rotation + translation parameters in Claessen). It would have been obvious to one of ordinary skill to incorporate Claessen's 3-D translation vector output into the modified invention of Feng in view of Azernikov such that the trained network not only outputs an axis vector for the correct tooth pose but also outputs a 3-D translation vector that places the origin of the correct tooth pose for the first segmented tooth representation.
Regarding Claim 8 the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination does not explicitly disclose but Claessen teaches converting the first segmented tooth representation in a point cloud representation (Claessen, Paragraph [0077], "a 3D deep neural network may be used that is capable of determining the canonical pose of optical scan data (a 3D point cloud) directly based on the point cloud data."; [0077], "These MLP deep neural networks are capable of directly processing points of a point cloud. Such neural networks may be trained to determine the canonical pose information based on the optical scan data as described in this application”;; it is noted 3D dental object directly from a point cloud representation of the 3D dental object
Feng provided a way of inviting alternative per-tooth representations; and Claessen supplies a concrete, well-known alternative: a point-cloud representation that PointNet-family MLPs can directly consume. It would have been obvious to one of ordinary skill in the art to convert Feng's per-tooth mesh into a point-cloud representation prior to feeding it to the trained neural network, particularly when that network is a PointNet, which is specifically engineered for point-cloud input. The motivation is interlocking once PointNet is selected as the trained neural network, a point-cloud per-tooth input is the natural, predictable representation for that network.
Regarding Claim 11, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches transforming the first segmented tooth representation [[using the initial tooth pose for the first segmented tooth representation comprises rotating the first segmented tooth representation ]] according to a negative value of coordinates of the origin of the initial tooth pose of the first segmented tooth representation (Feng, Paragraph [0048], "a mean value of coordinates of the vertices of the first digital data set may be calculated, to obtain coordinates of a center point, and then the coordinates of the center point is subtracted from the coordinates of the vertices of the first digital data set to obtain the second digital data set."; it is noted subtracts the origin/center-point coordinates from every vertex, which is exactly translating by the negative of the origin coordinates).
But the combination does not explicitly disclose rotating the first segmented tooth representation according to the initial tooth pose.
However, Claessen teaches using the initial tooth pose for the first segmented tooth representation comprises rotating the first segmented tooth representation (Claessen, Paragraph [0093] (re Fig. 3), "Using a trained 3D deep neural network, the 3D object may be (spatially) ‘normalized’ (i.e. re-oriented, re-positioned and scaled) 308 and defined based on an (orthogonal) canonical coordinate system. In the canonical coordinate system (x′,y′,z′) 306, the normalized 3D object 305 may have a canonical pose, in which specific features of the 3D object may be aligned with the axis of the canonical coordinate system."; [0104], "This data (both the 3D image data and the representation of the canonical system) are already downscaled appropriately as would result from processor 204. These data may then be rotated (e.g. employing a linear or other interpolation method) yielding 3D data as shown").
Claessen and Feng are analogous — both pre-process per-tooth 3D data into a canonical/normalized form to support neural-network inference. Feng supplies the explicit "translate by negative-of-origin coordinates"; Claessen supplies the complementary "re-orient (rotate) according to the canonical (initial) coordinate system". Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Claessen's re-orient/rotate normalization step into Feng's existing centralization step (in view of Azernikov) such that transforming the first segmented tooth representation using the initial tooth pose comprises both rotating the tooth according to the initial pose and translating the tooth by the negative of the origin coordinates. The motivation is that the initial tooth pose is anchored at the first tooth's centroid and oriented by the inter-centroid vector to the neighboring tooth's centroid, the per-tooth input to the trained neural network is most informative and the network easiest to train — when the tooth is presented in its initial-pose frame: rotated to align with the initial axes and translated so the centroid is at the origin.
Claim(s) 6 is/are rejected under 35 U.S.C. 103 as being unpatentable over Feng et al. (US 20210217233 A1, hereinafter Feng) in view of Azernikov et al. (US 20220215531 A1, hereinafter Azernikov) as applied to Claim 1 above and further in view of Sporbert et al. (US 20060263739 A1, hereinafter Sporbert).
Regarding Claim 6, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches that the output of the trained neural network is a three-component vector related to the [[ rotation ]] vector in axis of the per-tooth local coordinate system (Feng, Paragraph [0067], "The first predicted vector output by the MLP may be three components along the X-axis, Y-axis and Z-axis of the world coordinate system. When the label item is the z-axis of the local coordinate system, the first predicted vector is the z-axis of the local coordinate system predicted by the MLP.")
But the combination does not explicitly disclose [[ three-component vector related to the ]] rotation [[ vector in axis ]].
However, Claessen teaches wherein the output of the trained neural network comprises a 3-dimensional rotation vector [[ in axis-angle ]] per-tooth representation (Claessen, Paragraph [0156], "System 1100 may be employed to directly predict transformation parameters, e.g. applicable 3D rotations, 3D translations and 3D scaling defining how one received 3D image data set may be aligned to the other."; [0171], "with the application of the transformation parameters following from the system component, 1204 has been aligned or superimposed by means of 3D rotation and 3D translation.").
But the combination does not explicitly disclose [[ three-component vector related to the ]] rotation [[ vector in axis ]].
However, Sporbert teaches the axis-angle representation of a 3-D rotation applied to a tooth model (Sporbert, Paragraph [0104], " A 3×3 rotation matrix Mrot(v,α) is defined so that for xεR3 Mrot(v,α)·x is x rotated around the axis v by the angle α per Eq. (1)”; [0106], "where, v must have length 1 (one), i.e. √{square root over (vx 2+vy 2+vz 2)}=1."; [0170], "Apply rotation to our transformation matrix MT: MT←Mrot(v,α)·MT"; it is noted parameterizes a 3-D rotation by (i) a unit 3-D axis vector v = (vx, vy, vz) of length 1, and (ii) an angle α about that axis; the rotation is fully specified by this axis-and-angle pair. A 3-D rotation vector in axis-angle representation is, by definition, a 3-D vector whose direction is the rotation axis v and whose magnitude encodes the rotation angle α — exactly the (v, α) pair uses to parameterize a 3-D rotation applied to a tooth”).
Sporbert and Feng are analogous since both apply 3-D rotations to per-tooth 3-D models in a patient's dentition workflow; Sporbert establishes the axis-angle parameterization (axis v + angle α) as the form in which a 3-D rotation acting on a tooth is expressed mathematically. Claessen establishes that a trained 3-D deep neural network can directly output 3-D rotation parameters for a per-tooth representation. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Sporbert's axis-angle representation of a 3-D rotation with Feng's per-tooth-pose neural-network output (as already extended by Claessen to include rotation parameters) such that the network's rotation-related output is encoded as a 3-dimensional rotation vector in axis-angle representation — i.e., a 3-D vector whose direction is the rotation axis and whose magnitude corresponds to the rotation angle.
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Feng et al. (US 20210217233 A1, hereinafter Feng) in view of Azernikov et al. (US 20220215531 A1, hereinafter Azernikov) as applied to Claim 1 above and further in view of AnssariMoin et al (US 20210322136 A1, hereinafter AnssariMoin).
Regarding Claim 9, the combination of Feng and Azernikov teaches the invention in Claim 1.
The combination further teaches wherein the output of the trained neural network provides [[ translation and rotation parameters which, ]] when applied to the first segmented tooth representation, provide the correct tooth pose for the first segmented tooth representation (Feng, Paragraph [0115], "the y-axis and z-axis of the local coordinate system are determined based on the determined x-axis of the local coordinate system and the first predicted vector.") and (Feng, Paragraph [0067], "The first predicted vector output by the MLP may be three components along the X-axis, Y-axis and Z-axis");
But the combination does not explicitly disclose [[ the output of the trained neural network provides ]] translation and rotation parameters.
However, AnssariMoin teaches wherein the output of the trained neural network provides translation and rotation parameters which, when applied to the first segmented tooth representation, provide the correct tooth pose for the first segmented tooth representation (AnssariMoin, Paragraph [0021], "One or more of said plurality of training dental computed tomography scans may be associated with an indicator indicating an achieved transformation per tooth and/or an indicator indicating an attachment type per tooth, said transformation comprising a translation and/or a rotation per tooth (e.g. a transformation matrix or a vector)"; "The indicator indicating the achieved transformation per tooth allows the deep neural network to determine a transformation per tooth for a patient dental computed tomography scan and allows the desired final position per tooth to be determined based on this determined transformation."; [0080], "an indicator indicating an achieved transformation per tooth, e.g. a rotation and/or translation per tooth").
AnssariMoin and Feng are analogous since both train deep neural networks on per-tooth 3D dental data to produce per-tooth geometric transformation outputs. Feng supplies the per-tooth pose-determination framework with a network outputting rotation-related axis vectors. AnssariMoin supplies the complementary teaching that the same per-tooth network can jointly output translation and rotation parameters that, when applied to the per-tooth representation, yield the desired per-tooth pose. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate AnssariMoin's joint translation and rotation per-tooth network output into the modified invention of Feng in view of Azernikov such that the trained neural network outputs translation and rotation parameters which, when applied to the first segmented tooth representation, provide the correct tooth pose for the first segmented tooth representation. The motivation is that the claimed tooth pose by its own definition has both an origin (a translation) and three axes (a rotation), so a network that outputs both parameter types in one pass which completes the pose end-to-end with one inference, removing the need for downstream geometric reconstruction and matching Feng's existing per-tooth network-output paradigm.
Claim(s) 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Feng et al. (US 20210217233 A1, hereinafter Feng) in view of Azernikov et al. (US 20220215531 A1, hereinafter Azernikov) as applied to Claim 1 above and further in view of See et al. (US 10335251 B2, hereinafter See).
Regarding Claim 10, the combination further teaches displaying a [[ virtual ]] 3D representation on a display screen (Azernikov, Paragraph [0092], “FIG. 7 is for illustrative purposes only. In some embodiments, the number of columns and rows desired can be set in a configuration file or through a graphical user interface element such as an input filed for both values”);
But the combination does not explicitly disclose display screen with the correct tooth pose for the first segmented tooth representation.
However, See teaches displaying the virtual 3D representation on a display screen with the correct tooth pose for the first segmented tooth representation (See, Column 27, Line 5-8, "A) A method of defining a coordinate system for a virtual tooth comprising: displaying the tooth to a user in a three-dimensional viewing environment;"; Column 2, Line 45-47, "presenting a rendering of the surface model data in a user interface; receiving input data specifying a point on the rendered surface model associated with a tooth").
See and Feng are analogous since both define a per-tooth coordinate system (tooth pose) on a virtual 3D tooth representation derived from a patient's dentition. Feng determines the tooth pose computationally; See establishes that displaying that virtual 3D tooth representation on a display screen, together with its defined coordinate system, is a conventional and routinely useful step for orthodontic / restorative workflows. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate See's display step into the modified invention of Feng in view of Azernikov such that after the trained neural network outputs the correct tooth pose for the first segmented tooth representation, the virtual 3D representation is displayed on a display screen together with the correct tooth pose for that tooth. The motivation is the well-recognized clinical and user-interface need to verify and adjust the computed pose visually.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 11379975 B2 Classification and 3D modelling of 3D dento-maxillofacial structures using deep learning methods
US 11568533 B2 Automated classification and taxonomy of 3D teeth data using deep learning methods
US 20220398738 A1 METHOD FOR AUTOMATED TOOTH SEGMENTATION OF THREE DIMENSIONAL SCAN DATA USING DEEP LEARNING AND COMPUTER READABLE MEDIUM HAVING PROGRAM FOR PERFORMING THE METHOD
Any inquiry concerning this communication or earlier communications from the examiner should be directed to YUJANG TSWEI whose telephone number is (571)272-6669. The examiner can normally be reached 8:30am-5:30pm EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kent Chang can be reached on (571) 272-7667. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/YuJang Tswei/Primary Examiner, Art Unit 2614
Prosecution Insights